Guideline development

Purpose of guideline

The National Clinical Guideline for Stroke provides authoritative, evidence-based practice guidance to improve the quality of care delivered to every adult who has a stroke in the United Kingdom and Ireland, regardless of age, gender, type of stroke, location, or any other feature. The guideline is intended for:

• those providing care – nurses, doctors, therapists, care staff;
• those receiving care – patients, their families, their carers;
• those commissioning, providing or sanctioning stroke services;
• anyone seeking to improve the care of people with stroke. [2023]

Parties responsible for guideline

The 2023 edition, like the previous editions of this guideline, was produced by the Intercollegiate Stroke Working Party.  The Working Party was originally set up by the Royal College of Physicians and is a group of senior representatives from the professional bodies in England, Wales and Northern Ireland involved in stroke care, as well as the voluntary sector and patient voice representatives.  For this edition encompassing the whole of the UK and Ireland, the Working Party established a Guideline Development Group with representatives from the Working Party, from the Scottish Intercollegiate Guidelines Network (SIGN) and from the Irish National Clinical Programme for Stroke.  Production of the 2023 edition was supported by a project team based at the Sentinel Stroke National Audit Programme (SSNAP), King’s College London.  More information about contributors to the guideline is in Section 1.9 Contributors. [2023]

The 2023 guideline is endorsed for use in clinical practice by the Royal College of Physicians of London, the Scottish Intercollegiate Guidelines Network (SIGN) and the Royal College of Physicians of Ireland.  [2023]

Relationship to guidance from NICE and SIGN

In appraising the updated evidence base the Guideline Development Group has paid close attention to the content of the most recent NICE guidance on stroke:

• NICE guideline [NG128] Stroke and transient ischaemic attack in over 16s: diagnosis and initial management (last updated in 2022) https://www.nice.org.uk/guidance/ng128;
• NICE quality standard [QS2] Stroke in adults (last updated in 2016) https://www.nice.org.uk/guidance/qs2/chapter/Quality-statement-2-Intensity-of-stroke-rehabilitation;
• NICE guideline [CG162] Stroke rehabilitation in adults (last updated in 2013, update due late 2023) https://www.nice.org.uk/guidance/cg162. [2023]

This guideline is referred to as SIGN 166 and replaces the following SIGN guidelines:

• SIGN 119: Management of patients with stroke: Identification and management of dysphagia (published 2010, withdrawn 2020);
• SIGN 118: Management of patients with stroke: Rehabilitation, prevention and management of complications, and discharge planning (published 2010, withdrawn 2020);
• SIGN 108: Management of patients with stroke or TIA: Assessment, investigation, immediate management (published 2008, withdrawn 2018). [2023]

This guideline covers the management in adults (i.e. people aged over 16 years) of:

• stroke (ischaemic stroke and primary intracerebral haemorrhage [ICH]) and transient ischaemic attack (TIA), including ocular or retinal stroke and amaurosis fugax;
• subarachnoid haemorrhage (SAH): immediate management required at an admitting hospital. [2016]

The guideline does not cover:

• primary prevention of stroke;
• detailed recommendations on (neuro-)surgical techniques (but the role of surgery is addressed);
• management of children with stroke;
• surgical or neuroradiological interventions for SAH;;
• general aspects of healthcare, unless there are specific issues relating to stroke. [2016]

2023 edition

The 2023 edition is a partial update of the 2016 edition.  The scope of the update is available here.  This edition includes updated evidence published since 2015, with literature searches completed up to September 2022 and with some major publications since that date also included. [2023]

This 2023 edition of the guideline includes Scotland and the Republic of Ireland for the first time, as well as England, Wales, Northern Ireland, and the Isle of Man. This expansion of the guideline’s remit will affect some of the nomenclature, which differs between nations and health systems, and clinicians may need to interpret recommendations in the context of their healthcare system. [2023]

This guideline relates to those aspects of clinical management that are specific to stroke; it does not seek to address areas of routine clinical practice and good governance such as courtesy and respect for the individual, shared decision making and supporting patient choice, accurate record keeping etc.  [2016]

This guideline is not intended to overrule regulations or standards concerning the provision of services and should be considered in conjunction with them. In considering and implementing this guideline, users are advised to also consult and follow all appropriate legislation, standards and good practice. [2023]

No clinical guideline can account for every eventuality, and recommendations should be taken as statements that inform and guide the clinician, the patient and any other user, and not as rigid rules.  The clinician remains responsible for interpreting the recommendations taking into account the circumstances at hand (e.g. medical, psychosocial, cultural) and   including competing priorities for the patient and co-morbidities such as frailty, and for considering whether new evidence might exist that could alter the recommendation.  In doing so, clinicians should consider Sweeney’s three levels of significance when applying the evidence to the person in front of them: statistical significance (is the evidence valid?), clinical significance (does the evidence apply to this clinical situation?) and personal significance (does the evidence apply to this person’s particular circumstances and priorities?) (Sweeney et al, 1998).  Clinicians can reasonably expect guidelines to be unambiguous about the first and to give guidance about the second, but the third level of significance can only be understood within the relationship between the treating clinician and their patient, and may provide the justification for deviations from recommended management in particular cases. [2016]

This guideline has used several models or frameworks to structure its recommendations and layout.  In summary these are:

• the Donabedian model (Donabedian, 1978) for considering healthcare: structure, process and outcome;
• the healthcare process: diagnosis, assessment, intervention (treatment and support), and evaluation;
• the WHO international classification of functioning, disability and health (WHO ICF) model (World Health Organization, 1978; Wade & Halligan, 2004);
• time: prevention, acute, post-acute/recovery and long-term. [2016]

The WHO ICF model is a useful conceptual framework for disease management, particularly one with such long-term impact as stroke.  This framework is articulated in terms of:

• pathology (the disease processes within organs);
• impairment (symptoms/signs; the manifestations of disease in the individual);
• activities (the impact of impairments on the person’s usual activities);
• participation (the impact of activity limitations on a person’s place in family and society). [2016]

The same methodology was followed for the 2023 update as for the 2016 edition.  It includes seven distinct steps to ensure a thorough and rigorous process.  The detailed methodology is available in the methodology overview here.  The seven steps are:

1. Development of scope (see process and final scope, including research questions, here);
2. Searching the scientific literature (see search strategies here);
3. Selection of studies for inclusion;
4. Assessment of the quality of the evidence (see evidence tables here);
5. Moving from evidence to recommendations;
6. Health economic considerations;
7. External peer review and public consultation (see peer review and public consultation document here, and peer review comments and responses in the peer review report here). [2023]

Grading of recommendations and consensus recommendations

As with the 2016 and earlier editions, the methodology followed has:

• not graded recommendations. In many clinical guidelines, recommendations are given a grade which derives entirely from the design of the studies providing the evidence.  Methodologically strong evidence for less important interventions gives the linked recommendation an apparently higher priority than an important recommendation where the evidence is weaker.  The strength depends solely upon the study design and ignores other important features of the evidence such as its plausibility, generalisability, and the absolute benefit to the total population of people with stroke; [2016]
• used quantitative and qualitative evidence where appropriate. In choosing between making a recommendation based on less than perfect evidence and making no recommendation at all, the guideline seeks to guide practice using the best available evidence; [2016]
• allowed for recommendations to be developed by consensus:
  • in areas of practice where evidence is absent or of such poor quality or quantity that a recommendation cannot be derived; [2016]
  • where formal literature searching of a narrowly defined research question would not have adequately encompassed the clinical implications of the topic. (See the methodology guide here for more details). [2023]

Strength of recommendations

Depending on the strength of the evidence, recommendations in this guideline are either strong (a treatment or service ‘should be provided/offered’) or conditional (a treatment ‘should be considered’ or ‘may be considered’).  More detail about the wording of recommendations can be found in the methodology overview here. [2023]

NICE accreditation

The 2016 edition was accredited by NICE.  NICE has agreed to extend the accreditation to the 2023 edition.  This applies until December 2023.  NICE accreditation processes are changing and the accreditation status of this guideline may change in future. [2023]

Funding for the guideline project team was received from the following external sources on condition that the Guideline Development Group retained complete editorial independence over the guideline development process and content:

• Stroke Association;
• Johnson & Johnson Medical Ltd (through an educational grant agreement);
• Welsh Assembly Government;
• NIMAST (Northern Ireland Multidisciplinary Association for Stroke Teams). [2023]

The editors and the Intercollegiate Stroke Working Party wish to express their immense gratitude and appreciation for the financial support granted freely by these organisations, without which production of this edition would not have been possible. [2023]

The policy on declaration of interests is here and the interests registered by guideline contributors are here. [2023]

Users of this guideline should apply the general rule that if an intervention covered by the scope of this guideline is not mentioned in this guideline, then it is not recommended for use, and commissioners and service planners are not obliged to obtain it for the populations they serve. [2016]

A small number of specific recommendations that patients should not be offered a treatment except ‘in the context of clinical trial’ have been included.  This has been done when there is already some research which leaves uncertainty about the benefits and harms, but there is insufficient evidence to either recommend an intervention, or to avoid its use. [2016]

Recommendations in this guideline about the use of specific medicines (and devices) do not specify whether the medicine is licensed or approved by the Medicines and Healthcare products Regulatory Agency (MHRA), or European Medicines Agency (EMA), for that particular use.  It is the responsibility of the individual clinician and their healthcare provider to decide whether to permit the unlicensed/off-label use of medication in their formulary, including by referring to the source data from the appropriate regulator.  The Guideline Development Group may have considered it appropriate to recommend medication which has not been licensed for specific situations (e.g. aspirin in acute ischaemic stroke) or is not available in certain jurisdictions (e.g. injectable lipid-lowering therapies in the Republic of Ireland). Additional advice on the use of new and existing medicines is provided by the National Institute for Health and Care Excellence (NICE), which evaluates technologies for the NHS in England and Wales, and the Scottish Medicines Consortium (SMC) for NHSScotland and the National Centre for Pharmacoeconomics (NCPE) for the Health Service Executive (HSE) in Ireland. [2023]

The Intercollegiate Stroke Working Party is extremely grateful to the following groups who contributed their time and expertise (considerably above and beyond the usual call of duty) to producing this edition:

• the four guideline editors;
• members of the Guideline Development Group;
• topic group leads;
• members of topic groups and other contributors. [2023]

In addition, the Working Party is grateful to all those who contributed to their organisation’s peer review of the draft guideline and those who submitted public consultation comments.  Finally, the Working Party thanks the members of the stroke guideline project team and SIGN.  The guideline would not exist without the hard work of all the contributors, who are listed here. [2023]

Unchanged text from the 2016 edition is marked [2016].  Text is marked [2023] when:

• the recommendation and supporting text are new since the 2016 edition;
• the recommendation and supporting text have been updated since the 2016 edition as a result of new evidence;
• the recommendation was reviewed but not amended from the 2016 edition since the new evidence did not support a change. The supporting text has been updated. [2023]

The 2016 edition was prepared by the Intercollegiate Stroke Working Party (‘Working Party’).  The 2023 edition is the responsibility of the Guideline Development Group.  In the case of consensus recommendations, sources for 2016 recommendations refer to ‘Working Party consensus’, whereas sources for 2023 recommendations refer to ‘Guideline Development Group consensus’.  [2023]

Some 2016 text has been changed to ensure it is applicable to Scotland and Ireland as well as to England, Wales and Northern Ireland, without changing the sense. These changes are not marked [2023]. [2023]

This edition is published online only.  Apart from as indicated above, the text from 2016 is unchanged but may be in a different order from the 2016 publication.  Sections may also have been re-titled and re-numbered (particularly in the Rehabilitation and Recovery chapter). [2023]