Qualitative evidence synthesis to improve implementation of clinical guidelinesBMJ 2017; 356 doi: https://doi.org/10.1136/bmj.j80 (Published 16 January 2017) Cite this as: BMJ 2017;356:j80
- Christopher Carroll, reader in systematic review and evidence synthesis
As Sackett and colleagues wrote 20 years ago, evidence based practice involves the use of the “best external evidence” to inform clinical decision making.1 The published evidence used to underpin clinical guidelines, including those produced by the National Institute for Health and Care Excellence (NICE) in the UK, is almost exclusively quantitative. This is understandable as the principal focus is efficacy and safety: the aim is to establish what works. However, Sackett and colleagues were also clear that clinical practice should take account of patients’ preferences.
This is currently achieved by patient involvement in the process2 and by using primary qualitative research, which uses techniques such as interviews to explore how and why patients make the decisions they do.34 But a synthesis of such qualitative research studies paints a rich, subtle, and useful picture of patients’ experience, views, beliefs, and priorities, and could improve the implementation of clinical guidelines.
What is qualitative evidence synthesis?
Synthesis of quantitative studies using techniques such as meta-analysis promises greater power, more precise results, and the possibility of generalising from statistically representative samples.5 The basic rationale behind the synthesis of qualitative evidence is similar: to make the most of relevant studies for the purposes of policy and practice.4 The synthesis of several relevant qualitative studies can offer multiple perspectives as well as providing evidence of contradictory viewpoints that might otherwise be missed when considering a single study alone.4 Qualitative evidence synthesis also enables researchers to “go beyond” the findings of such primary research studies and produce something that is more than their simple sum.6
Because of its small sample sizes qualitative research is often criticised for lack of generalisability. However, this assumes that qualitative studies have the same purpose and measure the same outcomes as their quantitative equivalents. They do not. Rather, they are filling some of the gaps left by the quantitative evidence. The type of generalisability they offer is different. For example, qualitative study samples can be “informationally” rather than statistically representative, in the sense that they can offer information that is applicable to many other people with a similar condition or receiving similar treatments.4 Rather than seeking to offer an alternative method of measuring efficacy and safety, qualitative evidence and its synthesis is mainly aiming to provide something that the quantitative evidence often does not, such as identifying and explaining patient behaviours. It could be argued that it is essential for true evidence based practice.
Standards already exist for the conduct and reporting of qualitative evidence synthesis.7 Approaches to synthesis can be aggregative (such as narrative or framework synthesis or meta-aggregation, which summarise studies’ findings), interpretive (such as meta-ethnography or critical interpretive synthesis, which seek to generate completely original conceptualisations and theories based on the evidence), or a combination of the two (such as thematic synthesis).68 Framework, narrative, and thematic synthesis are particularly useful for answering questions about the uptake of interventions and for integrating quantitative and qualitative findings.689 These methods are therefore potentially the most appropriate for use in developing clinical guidelines. In the UK, NICE public health guidance already often uses a form of thematic synthesis and integrates quantitative and qualitative evidence using a narrative approach.10 Below, I will show how it could also be useful in clinical guidelines.
Enhanced clinical guidelines
Qualitative evidence synthesis has several potential benefits for clinical guidelines,4 but I will focus on patient preferences, in particular shared decision making or the principle of “nothing about me without me.”11 This principle requires that clinical decisions be consistent with the elicited preferences and values of the patient. It is an end in itself.11 Failure to take account of a patient’s needs and views contributes to lower levels of adherence to treatments and poorer clinical outcomes,12 whereas well conducted shared decision making improves patient satisfaction and willingness to follow treatment plans.13 These are key outcomes for any policy maker who wants to see research having its intended effect in practice.
Although NICE has a quality standard and clinical guideline on involving patients and, where appropriate, their family or other representatives in treatment decisions,1415 this guidance is quite generic. By contrast, a qualitative evidence synthesis of relevant studies can provide specific information about the many issues that need to be taken into account during shared decision making with particular groups of patients. This type of synthesis can therefore potentially offer a valuable supplement to the experiences of patient representatives on guideline panels, as the recent update of NICE guidelines for stroke rehabilitation show.16
Long term management of stroke
The full guideline17 identified a relevant synthesis of qualitative and quantitative evidence18 but essentially contained its own thematic synthesis of 17 qualitative studies, identifying relevant themes with supporting evidence listed. This published evidence suggested that patients and family members thought that health professionals viewed goal setting as relatively unimportant and as solely the professionals’ responsibility: decision making on this aspect of care was not being shared.
Together with the relevant quantitative evidence, these findings informed a series of evidence statements (section 6.2.317), which in turn informed the recommendations (6.2.5). These recommendations then appeared in the final guideline (CG162),16 which required that goal setting be conducted at specific meetings and be meaningful, relevant, challenging but achievable, time sensitive (reviewed regularly), and involve input from patients and their family or carers (box 1). The influence of the qualitative evidence and its synthesis is quite clear: the quantitative evidence only noted that standard procedures were not conducive to shared decision making; the qualitative evidence emphasised the importance of shared decision making and the specifics of how it should be achieved, and these were integrated in detail into the recommendations.
Box 1: Incorporation of qualitative evidence synthesis in NICE guidelines on long term management of stroke1617
Evidence from qualitative and quantitative studies (section 6.2.1)
Inhibitory factors such as limited time, presiding professional routines and the single opportunity to meet clinicians post discharge for secondary risk management (three qualitative studies: low to moderate confidence in studies)
Standard goal setting meeting, which is held away from the patient and with standard documentation, is not conducive to patient centred goal setting (quantitative study: low to moderate confidence)
Summary of challenges to patient participation in goal setting (6.2.3)17
Five studies highlighted factors inhibiting patients from participating in goal settings. These factors include: limited time, presiding professional routines, goal setting meeting which is held away from the patient, single opportunity to meet clinicians post discharge for secondary risk management, stroke pathology with its highly unpredictable recovery prognosis and its effects such as aphasia
Translation to clinical guideline16
1.2.8 Ensure that people with stroke have goals for their rehabilitation that:
Are meaningful and relevant to them
Focus on activity and participation
Are challenging but achievable
Include both short and long term elements
1.2.9 Ensure that goal setting meetings during stroke rehabilitation:
Are timetabled into the working week
Include the person with stroke and, where appropriate, their family or carer in the discussion
1.2.10 Ensure that during goal setting meetings, people with stroke are provided with:
An explanation of the goal setting process
The information they need in a format that is accessible to them
The support they need to make decisions and take an active part in setting goals
Unfortunately, such practice is rare in NICE clinical guidelines. Below, I use the examples of diabetes and cardiac rehabilitation to show how guidelines could be improved by including qualitative evidence synthesis.
Type 2 diabetes
The section on diet in the recently published NICE clinical guideline on type 2 diabetes (NG28) repeatedly recommends that health professionals, “provide … advice,” “emphasise advice,” and “discourage” or “encourage” certain actions.19 Strategies therefore emanate from the relevant health professional alone. Yet a recent qualitative evidence synthesis of 37 studies emphasised the importance of shared decision making because patients with type 2 diabetes and their families felt that communication with health professionals was often difficult and their opinions were not acknowledged.20 The evidence synthesis indicated that a shift from “advice” to negotiation was needed; effort should be made to elicit the concerns, needs, and preferences of patients and their families. Box 2 gives a possible revised recommendation for the section on nutritional advice.
Box 2: How NICE recommendations for type 2 diabetes19 could be enhanced by findings from qualitative evidence synthesis20
Current guideline recommendation
Provide individualised and ongoing nutritional advice …
Provide dietary advice in a form sensitive to the person’s needs, culture and beliefs …
Emphasise advice on healthy balanced eating that is applicable to the general population …
Themes from qualitative evidence synthesis
Difficulty communicating with healthcare provider—Individual has difficulty communicating needs, questions, and concerns with healthcare provider
Respectful communication—Transferring of information in a way that is understood by the sender and receiver with consideration for feelings, rights, wishes, or traditions. It is the acknowledgment that both parties and their opinions have value
Possible enhanced recommendation
Ensure that the person with type 2 diabetes:
Is given an explanation of why healthy balanced eating is important
Is given an explanation of how an agreed dietary plan will have benefits for them
Ensure that meetings regarding nutrition are ongoing and individualised, and the person with type 2 diabetes is given the information and support they need to make decisions and take an active part in identifying the best diet for them and that the information is in a format that is accessible to them
Ensure that the person with type 2 diabetes
Is asked about their concerns and needs
Is asked about any restrictions to their diet governed by their feelings, rights, wishes, religion, or traditions
Participates in deciding what dietary changes are appropriate and achievable for them
Long term cardiac rehabilitation
As with type 2 diabetes, evidence on patients’ views and experiences is limited in the NICE clinical guideline on cardiac rehabilitation (CG172).21 Two of the key findings from a qualitative evidence synthesis of 90 studies looking at patients’ views of cardiac rehabilitation are: patients’ sense of lacking any control over their condition and being unconvinced that the interventions on offer would produce positive outcomes.22
Several of the included studies reported that some cardiac rehabilitation patients focus only on the avoidance of stress to reduce the chance of another heart attack (for example) rather than modifying diet, physical activity, or smoking—three key points in the guidance (CG 172). Given that attendance of rehabilitation programmes is a known problem,22 a decision making process that explicitly addresses patients’ potential mindsets might lead to greater patient satisfaction with agreed treatment plans and improved clinical outcomes. The guideline can be made more specific (box 3).
Box 3: Using qualitative evidence to enhance NICE recommendations for cardiac rehabilitation21
Current guideline recommendation on encouraging people to attend
Establish people’s health beliefs and their specific illness perceptions before offering appropriate lifestyle advice and to encourage attendance at a cardiac rehabilitation programme.
Theme from qualitative evidence synthesis
Patients perceived heart disease as defying any attempts to reduce risk. For example, risk of acute myocardial infarction was perceived to be unpredictable, inevitable, and uncontrollable, irrespective of whether the underlying heart condition was seen as low or high severity. Likewise, participants expressed a low sense of control over their future health.
Possible enhanced recommendation
Involve patients in decision making by establishing their health beliefs and specific illness perceptions:
Ask patients why they think they had a heart attack
Ask patients whether they think their condition can be controlled
Explain the relation between lifestyle and heart disease
Explain how appropriate lifestyle behaviours and attendance at a cardiac rehabilitation programme can give the patient greater control over their condition
Agree appropriate and achievable lifestyle changes and rehabilitation programme attendance that is meaningful to the patient
Putting evidence into practice
Clinical guidelines and quality standards might stress the need for decision making to be shared, but it is the synthesis of qualitative evidence that details what this negotiation should involve for any particular condition and its treatment.14 By accessing and using evidence on patients’ anxieties, beliefs, and preferences, which can be highly condition specific, recommendations in clinical guidelines can be tailored and enhanced. Treatment plans might have a greater chance of being followed if they are the result of a negotiation that seeks to cover and address topics of known importance to particular patient groups.13
However, the use of this evidence is not without problems.4 Although there are many hundreds of published qualitative evidence syntheses that can be used for clinical guidelines, they might not be available for every indication. Fortunately, there are pragmatic and relatively rapid methods of qualitative evidence synthesis that guideline developers could use to fill that gap.9 Also, generic qualitative evidence synthesis reporting guidelines exist,7 others are being developed for particular methods,23 and standards are evolving to establish the level of confidence users can ascribe to the findings of such syntheses.24
Despite the availability of methods for integrating quantitative and qualitative evidence,68 there is no ready made toolkit for doing so. The NICE stroke guideline and public health programme10 both offer relevant templates, but future work should seek to identify the most appropriate approach for clinical guidelines. The qualitative evidence might also come from settings that are not directly applicable to the NHS, so this needs to be taken into account, though the same problem can apply to quantitative evidence. Nevertheless, as the examples described above suggest, such evidence, carefully considered and integrated with the quantitative evidence, can offer a highly useful addition to the expert and patient opinion currently used in the guideline development process.
Simply recommending the general principle of shared decision making in clinical guidelines does not mean issues important to patients will be addressed
Qualitative evidence synthesis can help guideline developers identify these issues and include specific recommendations
By making use of this evidence, clinical guidelines can be more informed, richer, and context specific
This has potential benefits for patient satisfaction and clinical outcomes
Contributors and sources: CC is a member of the Sheffield Technology Assessment Group, conducting systematic reviews for NICE, and the codeveloper of the qualitative evidence synthesis method, “best fit” framework synthesis. He is a member of the NICE Interventional Procedures Advisory Committee.
Competing interests: I have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.
Provenance and peer review: Not commissioned; externally peer reviewed.