Before presenting different enablers for SDM in guidelines, we will discuss whether or not SDM is more applicable for some recommendations than others. We will also discuss whether the strength of a recommendation has a role in deciding this.

Some models that assess the potential for SDM in guidelines suggest that weak recommendations are most appropriate for sharing decisions. This applies especially to the GRADE framework: ‘When a recommendation is weak, clinicians and other health care providers need to devote more time to the process of shared decision-making by which they ensure that the informed choice reflects individual values and preferences.’ (Andrews et al. 2013)

However, in their DECIDE work, Fearns et al. found that weak recommendations triggered strong negative reactions from members of the public (2016). Although they understood that a weak recommendation was less strongly endorsed, they often interpreted it as the intervention not being effective.

Weak recommendations are made when there are different options, including no intervention, that are equally sensible and choices may differ largely among individual patients depending on their individual situations. Therefore, without any question, SDM applies here. However, SDM may be equally important in relation to strong recommendations, if the situation is appropriate (that is, it is not an emergency situation). It is assumed for strong recommendations that most informed people would decide in favour of the recommendation. But this raises the question of who makes that assumption. The experience, concerns and preferences of the guideline developers may not be shared by all patients. Several people may decide differently, in the context of their particular circumstances. These circumstances are characterised by the International classification of functioning, disability and health’s (ICF) contextual factors and may present good reasons for deciding against a well-established, evidence-based intervention (World Health Organization [WHO] 2001).

These contextual factors include:

• environmental factors: factors that are not within the person’s control, such as family, work, government agencies, laws, and cultural beliefs
• personal factors: factors such as race, gender, age, educational level, coping styles, health status, and risk attitudes, which vary widely among individuals and cultures.

Offering SDM enables patients to make a decision that best suits their individual and environmental conditions. Box 1 provides an example of a guideline panel deciding to provide a decision aid for a strong recommendation.

Box 1 Case study of decision support for a decision on taking statins for coronary artery disease in the German national disease management guideline on chronic coronary artery disease (Bundesärztekammer 2019)

Background High grade evidence for statins in patients with coronary artery disease (CAD) shows consistent effects on mortality and morbidity and only rare adverse events. However, myalgia has often been reported under statin use, although randomised control trials showed no difference between intervention and control groups therefore suggesting a nocebo effect. Most guidelines give a strong recommendation in favour of statins for patients with CAD. Rationale Based on clinical experience, the large multidisciplinary guideline panel identified that the decision to start or continue statins was a key situation for SDM, mainly for 2 reasons: • they assumed there was considerable overtreatment in older patients on polypharmacy, and • according to their clinical experience, some patients refuse to take statins because of the false belief that these caused muscle pain. Intervention A decision support tool was designed, based on the results of the systematic review for the guideline, that provided a drugs fact box for statins and additional plain language information, which clearly explained why myalgia was most probably not caused by the medication. It was formatted so that it could be printed and handed out to patients (as a short leaflet) as well as being used online. The tool underwent formal consensus and was integrated into and published with the guideline. It was provided to clinicians (through the guideline and the webpage for physicians), but was directly and freely accessible for patients (though the patient webpage). Evaluation showed that physicians thought such tools were helpful in the clinical encounter and that most would use them with their patients (Schwarz et al. 2019).

Preference sensitivity

It is widely recognised that SDM is a concept to tackle preference sensitive decisions (Elwyn et al. 2009). As indicated by the above ‘GRADE’ definition, a considerable group of guideline developers and methodologists assume that weak recommendations indicate such preference sensitive decisions. However, not all guideline groups are consistent in their choice of the grade of recommendation for identical clinical questions. Every recommendation is not only based on the underlying evidence but on the collective value judgement of a specific group, influenced by, among others, their experience, academic interests and professional background. Hence, although there is some overlap between preference sensitive situations and situations where guideline panels make weak recommendations, they are not totally congruent. The concept of preference sensitivity needs to be operationalised to help identify situations where SDM is most appropriate. An example of operationalisation is shown in box 2.

Box 2 Case study on operationalisation of ‘preference sensitivity’ to guide the development of decision support tools for guideline developers

The National Institute for Health and Care Excellence (NICE) uses an operationalised concept of preference sensitivity in its process guide for decision aids. The concept is used to determine whether or not a decision point in a guideline needs more support to enable SDM, and has characterised those decisions as follows: ‘Preference-sensitive decision points are points where the person’s values and preferences are particularly important. They occur when either: • There are 2 or more options for investigation, treatment or care that deliver similar outcomes but: : they have different types of harms and benefits which people may value differently, or : the likelihood of the harms or benefits may differ, or : the practicalities of the options are different (for example, the choice is between medicine and surgery, or the requirements for monitoring differ), or : people may consider the overall risks of harms for any of the options outweigh the overall benefits compared with no treatment or • The choice between an investigation, treatment or care option and the option of ‘no treatment’ is finely balanced.’ (NICE 2018c)

Strong recommendations against interventions – providing rationales, not options

Strong recommendations against interventions are rare and usually made when a panel is sure that an intervention provides no benefit but substantial harm. It is hard to imagine that any reasonable and informed person would choose an option that does them no good, but instead, puts them at risk of serious harm. SDM requires commitment from patients and for some it may even be hard work discussing options with their doctor, reading, understanding the evidence, weighing up the benefits and harms, and making a decision. It is reasonable to ask whether the commitment, time and resources are necessary when there is no sensible alternative, and if patients and HCPs might rather need another type of information in these situations, that is, information that explains and supports a recommendation instead of offering options. Evidence from DECIDE indicates that when making a decision, patients want an honest explanation of the rationale behind such a recommendation (Fearns et al. 2016). 

For example, imaging for low back pain is not recommended in the absence of red flags indicating a serious condition. Evidence has shown that imaging does not lead to better outcomes but may cause unnecessary treatment and increase the risk of the problem becoming chronic (Chou et al. 2011). Most guideline panels make a strong recommendation against imaging for low back pain. However, research indicates that patient expectation may be a driver of unnecessary testing because of the false belief that imaging is beneficial, and because patients may feel uncomfortable and not acknowledged when not getting a test or a treatment (Warner et al. 2016, Parmar 2016, Pathirana et al. 2017).

The idea that recommendations against interventions need transparent communication is reflected in the concept of ‘Choosing Wisely’. In the US, the Choosing Wisely programme is collaborating with Consumer Reports to provide patient leaflets for all Choosing Wisely recommendations. However, these recommendations seem to not have been implemented very well (Hong et al. 2012). They have also been shown to not adhere to international quality standards (Legaré et al. 2016). This indicates that information supporting negative recommendations:

• needs to be provided to patients and physicians at the same time, and
• has to be designed carefully and respect patients’ autonomy.

In the NICE guideline on dementia (NG97; NICE 2018a) a strong recommendation is made against routine enteral feeding for people living with severe dementia. A decision support tool on enteral (tube) feeding for people living with severe dementia was developed to guide discussions between HCPs and patients, their carers and relatives (NICE 2018b). Although being called a decision aid, the tool clearly explains the certainty of the evidence regarding harms and the lack of evidence for any benefit, thereby being strongly supportive of the recommendation.

In conclusion, we strongly suggest that:

• strong recommendations in favour of interventions (if not addressing emergency situations) be carefully considered for SDM with decision tools provided
• strong recommendations against interventions need careful discussion with the patient, supported by information that explains the rationale rather than offering options.