WSO - June 2023 - 522

522
International Journal of Stroke 18(5)
Quality of guidelines
The AGREE II tool demonstrated that guidelines generally
did not consider resource implications or implementation.
Thus, our subjective assessments of applicability had to be
made based on own knowledge of resources in different
health care settings. Ideally, services across the world
should consider the cost-benefit of interventions based on
local costs and resources available.
Strengths of our approach
We published our protocol on PROSPERO; we followed
this protocol except for the exclusion of pediatric guidelines
because we did not have sufficient resources to evaluate
those documents. We performed comprehensive
searches, we convened an international group of experts in
topic areas and systematic reviews with a balance of gender,
seniority, and location.
Weaknesses of our approach
To the best of our knowledge, there is no published methodology
to synthesize the results of multiple guidelines.
Although our approach was rigorous, some subjectivity,
based on the judgment of the group members, was needed
when guidelines differed in strength of recommendations
for particular interventions.
There is no systematic way to our knowledge to investigate
how the AGREE II tool can be used to identify which
domain(s) influences recommendations. The previous
WSO 2014 guidance took just two domains and selected 10
papers which had good scores in those two domains.
Instead, we considered 200 publications identified in our
searches. Further work is needed to develop ways to explore
the influence of guideline quality on recommendations
made. Researchers wishing to take this forward should contact
us for access to the AGREE II assessments.
Although all the group was medically qualified and
had broad range of experience from hyperacute care to
rehabilitation and life after stroke, we did not have experts
in other disciplines. We invited the WSO board members
(which include stroke support organizations, with PWLE)
to comment on our draft. We received several professional
responses and lay responses. The next WSO guideline
committee should seek to have more balanced professional/PWLE
membership. This may require funding
to enable PWLE to provide their time and expertise.
Also, PWLE will be important in the dissemination stage
of our work.
We did not search gray literature (i.e. literature that is not
formally published in sources such as books or journal articles)
or contact experts in the field to identify local/regional
guidelines-due to insufficient resource. All the guidelines
we reviewed were based on the same trials and systematic
International Journal of Stroke, 18(5)
reviews, so it is unlikely that gray literature/regional guidelines
would have come to different conclusions.
Ideally two authors would have extracted recommendations
and reviewed quality for all guidelines, but this was
not always feasible. It is unclear whether this would have
improved the quality of our review process or changed our
final recommendations.
Although we initially planned to exclude recommendation
on TIA, we decided to modify the original PROSPERO
protocol by including these recommendations. Reasons
were that most recommendations for the prevention of
ischemic stroke are also valid for TIAs and for specific recommendations,
it would have resulted in reporting recommendations
for stroke but not for TIA (e.g. dual antiplatelet
therapy for minor stroke or TIA).
How to use this information
This is the most comprehensive systematic evaluation of all
world guidelines in stroke; however, the existence of guideline
recommendations is meaningless without efforts to
implement them in clinical practice. Services will need to
consider the cost-benefits of different interventions in their
own settings. Implementation of evidence is often challenging
particularly for complex interventions that require
changes in service structure and processes, or interventions
for which new funding is needed. It is sometimes easier
when disinvestment is needed or when simple changes
are needed to organization of care. The recommendations
we found generally require additional resources or reorganization
of care pathways, although some of the recommendations
require disinvestment (e.g. avoid very early
mobilization). Based on the recommendations from this
review, we then developed performance metrics for
services, which may be used in audits (Table 4). These
are at service level, process level, or individual patient
level. A review of practice in line with findings in this
review should be undertaken at multiple levels: individual
clinicians, hospital teams/services, and across whole health
systems.
Implications for future guideline
development and research
Guideline development is a resource-intensive process. Not
only did we identify multiple superseded versions of guidelines,
but there was duplication of work among different
guideline groups, with the same trials and systematic
reviews being cited in multiple guidelines. A central repository
of stroke trials could be coordinated by WSO. Cochrane
Stroke has maintained such a database for many years
(Database of Research in Stroke (DORIS), www.askdoris.
org/m), but with the disbanding of the UK Cochrane groups,
DORIS will no longer be updated. However, with a modest
http://www.askdoris.org/m http://www.askdoris.org/m
WSO - June 2023

Table of Contents for the Digital Edition of WSO - June 2023
