Top 10 research priorities relating to life after stroke - consensus from stroke survivors, caregivers, and health professionals

Top 10 research priorities relating to life after stroke – consensus from strokesurvivors, caregivers, and health professionals

Alex Pollock1*, Bridget St George1, Mark Fenton2, and Lester Firkins3

Background Research resources should address the issues thatare most important to people affected by a particular health-care problem. Systematic identification of stroke survivor,caregiver, and health professional priorities would ensure thatscarce research resources are directed to areas that mattermost to people affected by stroke.Aims We aimed to identify the top 10 research priorities relat-ing to life after stroke, as agreed by stroke survivors, caregiv-ers, and health professionals.Methods Key stages involved establishing a priority settingpartnership; gathering treatment uncertainties from strokesurvivors, caregivers, and health professionals relating to lifeafter stroke (using surveys administered by e-mail, post, andat face-to-face meetings); checking submitted treatmentuncertainties to ensure that they were clear, unansweredquestions about the effects of a treatment/intervention;interim prioritization to identify the highest priority ques-tions (objectively identified from ranking of personal priori-ties by original survey respondents); and a final consensusmeeting to reach agreement on the top 10 research priorities.Results We gathered 548 research questions that wererefined into 226 unique unanswered treatment uncertainties.Ninety-seven respondents completed the interim prioritizationprocess, objectively identifying 24 shared priority treatmentuncertainties. A representative group of 28 stroke survivors,caregivers, and health professionals attended a final meeting,reaching consensus on the top 10 research priorities relatingto life after stroke.Six of the agreed top 10 research priorities related to specificstroke-related impairments, including cognition, aphasia,vision, upper limb, mobility, and fatigue. Three related to moresocial aspects of ‘living with stroke’ including coming to termswith long-term consequences, confidence, and helping strokesurvivors and their families ‘cope’ with speech problems. Onerelated to the secondary consequences of stroke and subse-quent stroke prevention.Conclusions The top 10 research priorities relating to life afterstroke have been identified using a rigorous and person-centered approach. These should be used to inform the priori-tization and funding of future research relating to life afterstroke.Key words: James Lind Alliance, life after stroke, prioritization,rehabilitation, research priority, stroke

Introduction

Involving patients and caregivers in setting research and funding

agendas improves the quality, relevance, implementation, and

cost-effectiveness of research conducted (1–3) and ensures that

resources are used to address the priorities of people affected by a

particular healthcare problem (4–6). Involving stroke survivors,

their families and caregivers in research prioritization is a key

objective of The Stroke Association in the United Kingdom

(UK) (7). However, there is evidence that research ideas valued by

stroke health professionals may not be considered worth pursuing

by stroke survivors and caregivers (8). Systematic identification of

stroke survivor and caregiver priorities would ensure that scarce

research resources are directed to areas that matter most to people

affected by stroke. This in turn would ensure that the product of

funded research is valued and relevant.

Within the UK, the James Lind Alliance (JLA) has been estab-

lished specifically to help achieve meaningful patient and clinician

involvement in research priority setting and to ensure that those

who conduct and fund health research are aware of what gaps

in knowledge matter most to patients, caregivers, and clinicians

(9–11). Patients, clinicians, and their representative groups come

together within equal ‘priority setting partnerships’ (PSPs), which

identify and prioritize unanswered questions (‘treatment uncer-

tainties’) (12). A treatment uncertainty is defined as a question

about the effects of a healthcare intervention (including preven-

tion, testing, and rehabilitation) for which there are no up-to-

date, reliable systematic reviews of research evidence or for which

there are up-to-date systematic reviews of research evidence con-

firming that uncertainty exists.

Priority setting methods have been established and used by a

number of JLA PSPs (e.g. 13–16). Key stages involve establishing

a priority setting partnership; eliciting the treatment uncertain-

ties of patients, caregivers, and clinicians; checking the submitted

uncertainties; an interim prioritization stage; and a consensus

meeting to agree the top 10 research priorities; this process is

summarized in Fig. 1. The methods adopted for the final priority

setting (consensus meeting) stage are based on the nominal

group technique, which is a well-established and well-described

approach to decision making (12,17). The overall aim is to achieve

equitable participation and contribution from representative

groups of patients, caregivers, and clinicians at all stages of this

process. The methods allow for a certain amount of flexibility, in

response to the individual needs and goals of the partner organi-

sations within the PSP (18). Despite this flexibility, JLA methods

strive to be methodologically defensible and aim to ensure

transparency, accountability, and fairness. The JLA philosophy

is pragmatic, with recognition that JLA methods are developing

Correspondence: Alex Pollock*, NMAHP Research Unit, BuchananHouse, Glasgow Caledonian University, Glasgow G4 0BA, Scotland, UK.E-mail: [email protected] Midwifery and Allied Health Professions (NMAHP) ResearchUnit, Glasgow Caledonian University, Glasgow, UK2UK Database of Uncertainties about the Effects of Treatments (UKDUETs), NHS Evidence, National Institute for Health and ClinicalExcellence, Manchester, UK3James Lind Alliance, James Lind Initiative, Summertown Pavilion,Oxford, UK

Conflicts of interest: none declared.

DOI: 10.1111/j.1747-4949.2012.00942.x

Research

© 2012 The Authors.International Journal of Stroke © 2012 World Stroke Organization

Vol ••, •• 2012, ••–•• 1

rather than established; reflection on and refinement of the

methods occurs as an essential component of each priority setting

partnership.

Aims

We established a JLA PSP, which aimed to:

• identify the top 10 research priorities relating to lifeafter stroke with equitable participation and contributionsfrom stroke survivors, caregivers, and health professionals;and

• inform those whom conduct and fund health research whatgaps in knowledge matter most to stroke survivors, caregivers,and health professionals working in stroke care.

For this priority setting project, ‘life after stroke’ was defined

as including all aspects of general management and rehabi-

litation after stroke; prevention and management of com-

plications arising from stroke; and life after stroke (including

return to work, family roles, leisure pursuits, relationships,

and quality of life). ‘Life after stroke’ had no restrictions relating

to length of time after stroke. For the purposes of this PSP,

the definition of ‘life after stroke’ did not include assessment,

investigation, immediate management, or pharmacological or

surgical interventions aimed at secondary prevention of stroke

or transient ischemic attack (TIA).

Methods

As part of a project to develop a web-based Database of

Research in Stroke (DORIS), providing easy access to existing

evidence, ongoing research, and research uncertainties and gaps

relating to stroke rehabilitation (19), we established a JLA PSP in

February 2010. Treatment uncertainties were gathered between

June and December 2010 and, following systematic and rigorous

checking, interim prioritization took place between July and

October 2011. The final priority setting consensus conference,

in which the top 10 priorities were agreed, took place in

November 2011.

The life after stroke PSP was managed by a steering group

comprising a stroke survivor, caregiver, nurse, physician, allied

health professional, researcher and representatives from key

national stroke charities/patient organizations, and from the JLA.

The steering group defined the scope of this partnership and

developed a protocol detailing the methods to be used. The defi-

nitions of scope were used to identify appropriate participants

and participant organizations. However, the steering group made

the conscious decision not to provide these definitions on any

documentation requesting contributions to the project. This was

to ensure that submissions to the process were not restricted in

any way and to enable participants to use their own interpretation

of the phrase ‘life after stroke’.

Fig. 1 Key stages of priority setting process (numbers of results in brackets).

Research A. Pollock et al.


2 Vol ••, •• 2012, ••–••

Gather treatment uncertaintiesThe primary aim of this phase was to gather treatment uncertain-

ties, facilitating representative and equitable engagement from

stroke survivors, caregivers, and health professionals throughout

Scotland.

Stroke patient and caregiver groups, stroke research and deliv-

ery of care organizations, and community, long-term care and key

health professional groups were identified. Information about the

project, including a request for submissions of treatment uncer-

tainties, was produced in collaboration with a stroke survivor and

caregiver, written in plain English and designed to be understood

by lay users. An aphasia-friendly information sheet was developed

in collaboration with a speech and language therapist and

someone with aphasia. An audio version of the presentation was

recorded, designed to make the project information accessible to

people with visual problems or reading difficulties. We accepted

submissions of treatment uncertainties in any format – electronic,

mail, or verbally (either face-to-face, by telephone, or via a com-

munication partner). These materials were disseminated using a

variety of strategies to reach and facilitate involvement of the

identified groups:

E-mailAll identified groups and organizations were e-mailed

project information and a request to submit their treatment

uncertainties.

MailProject information, response forms, and prepaid return enve-

lopes were posted to the membership of patient groups if this was

the routine mode of communication.

InternetThe audio version of the presentation was made available online.

Links to this presentation were made from relevant websites and

provided on all project materials.

Face-to-face meetingsIn order to facilitate involvement of stroke survivors and care-

givers throughout all geographical regions of Scotland and to

reach people who may not receive and/or respond to e-mail or

postal information, a member of the steering group visited exist-

ing stroke support groups/clubs in each of the 11 National Health

Service (NHS) geographical locations on the Scottish mainland,

and we arranged for information leaflets to be distributed at

groups meetings on all three NHS Scottish Island locations.

A standard presentation was developed for delivery at stroke

support groups/clubs, available in a number of formats

(PowerPoint presentation, table top presentation, descriptive

presentation).

Attending professional meetingsIn order to facilitate involvement of a range of health profession-

als working in stroke care, including those who may not receive or

respond to e-mail information, we attended a number of national

meetings and conferences.

In addition to gathering treatment uncertainties from stroke

survivors, caregivers, and health professionals, we identified and

included relevant published research questions and priorities. We

searched UK national guidelines and published prioritization

projects to identify relevant treatment uncertainties.

Check treatment uncertaintiesThis stage established the ‘unique unanswered treatment uncer-

tainties’ relating to life after stroke. All submitted uncertainties

were transcribed onto a management database. Where informa-

tion was available, we documented whether the submission was

from a stroke survivor, caregiver, health professional or other, and

whether the submission was from an individual or from a group/

organization.

Each submitted treatment uncertainty was checked to ensure

that it was a question about the effect of a treatment/intervention.

Nontreatment questions, for example those relating to local

services, prevalence, and prognosis, were excluded. All the

remaining treatment uncertainties were formatted using a

‘PICO’ (population-intervention-comparison-outcome) structure

(20,21), and then grouped under relevant descriptive headings.

Any duplicate questions were identified and merged into ‘indica-

tive’ questions. These processes were carried out by a researcher

(AP/BSG), and each individual decision checked independently

by a health professional, a stroke survivor, and a caregiver.

Where wording of original submitted uncertainties was

changed (i.e. formatted) or duplicate questions were merged,

we kept a record of all the original submissions relating to the

indicative question and who had submitted these.

For each of the remaining unique treatment uncertainties, we

systematically searched published literature to check that each

was a true uncertainty and was not already answered by research

evidence. We used advanced search techniques (where possible,

combining ‘stroke’ with terms for problem and intervention) to

search the Cochrane Library. If there was a Cochrane review

relevant to the treatment uncertainty, we viewed the implications

for research. If further research was recommended, then this was

classed as a true uncertainty. If there were no relevant Cochrane

reviews, we searched for other published systematic reviews and

repeated the above process as per Cochrane reviews. If there was

(i) no published systematic review; (ii) a published review with

a search more than two-years out of date; or (iii) a published

review that identified the need for further primary research, then it

was assumed that further research was necessary, and the question

was a true uncertainty. Each uncertainty was also checked against

evidence synthesised within the Scottish Intercollegiate Guidelines

Network (SIGN) stroke rehabilitation guidelines (22). We system-

atically documented the evidence considered and whether further

research was required or not for each unique treatment uncer-

tainty. All stages in the process were checked independently by a

health professional and a stroke survivor or caregiver.

Interim prioritizationThe aim of this stage was to objectively identify the shared top

priorities relating to life after stroke. The unique unanswered

treatment uncertainties were grouped under key headings and

distributed with a request for people to identify, and rank, their

personal 10 most important priorities from this list. This was sent

by post or e-mail to all individual stroke survivors, caregivers, and

health professionals who had submitted uncertainties and also

ResearchA. Pollock et al.


Vol ••, •• 2012, ••–•• 3

e-mailed to all the groups and organizations identified in the

previous stage. To facilitate equitable involvement of individuals

with barriers to participation (e.g. stroke survivors with commu-

nication or cognitive impairments), a researcher visited a number

of stroke support groups/clubs and provided face-to-face instruc-

tions and support with this process.

The 10 priorities returned by each respondent were entered

into a spreadsheet. We determined the combined ranked order of

the unanswered treatment uncertainties using two methods. First,

we created a combined rank order independent of the ranking

applied by individual respondents; and second, we created a

combined rank order which reflected the ranking applied by indi-

vidual respondents. We explored the combined ranked order of

treatment uncertainties generated by each of these methods, as

provided by stroke survivors/caregivers and health professionals.

The aim was to generate a list of no more than 25 shared priorities

representing the views of stroke survivors, caregivers, and health

professionals.

Final priority settingThis phase involved a meeting to reach consensus on the top 10

research priorities relating to life after stroke. Consensus meeting

participants were purposively selected to provide a representative

group of 20–30 stroke survivors, caregivers, and health profes-

sionals, approximately half of whom were stroke survivors or

caregivers (with a range of demographics including age, time

since stroke, severity of stroke, impairments, geographical

location) and half of whom were health professionals (represent-

ing key professions, with a range of experience, care setting, and

geographical location).

All consensus meeting participants were provided with mate-

rials at least two-weeks prior to the meeting, this included an

information leaflet and the list of shared priorities identified

during the interim prioritization process. Participants were asked

prior to the meeting, to rank in order of priority the importance

of these questions to them.

In the first phase of the consensus meeting participants were

divided into three groups with the type of participants balanced

across the groups. Each group had a facilitator with knowledge of

stroke but who was independent of this priority setting project

and the James Lind Alliance. The facilitators were briefed by

members of the JLA on the importance of ensuring equitable

participation of all group members. Each group was given a series

of A4 cards, each with one of the research questions printed

clearly on the front. On the back of each card was summarized

background information including the original submitted treat-

ment uncertainties, who had submitted the treatment uncertain-

ties (stroke survivors, caregivers, health professionals; individuals

or groups), and the results of the voting during the interim

prioritization process. Each group was asked to rank all of these

cards/research questions in order of importance. The facilitator

supported this process, but the method of gaining agreement on

the rank order of the questions was not prescribed.

For the second phase of the meeting, the ranked order from

each of the three groups was summed to give a total ‘score’, a

combined ranking was determined, and these results presented to

all the meeting participants within a plenary session. Questions

were highlighted that had been placed in the top 10 by one or

more of the three groups but not placed in the combined ranking

top 10. Participants were invited to discuss and comment on these

questions, with particular emphasis on the reasons for differences

in the rankings applied by the different groups. The purpose

of this discussion was to reach consensus on an overall top 10

ranking. To facilitate this process, the questions were written on

cards and were laid out in priority order in a central place in the

room. Group participants discussed and moved the priority order

of cards until consensus was reached. This process continued

until there was unanimous agreement on the placement of the top

10 research priorities.

Results

Gather treatment uncertaintiesProject information and response forms were e-mailed to

members of seven national stroke patient and caregiver groups;

four national stroke groups/research organizations; five national

community care/long-term care organizations; and 17 health pro-

fessional organizations. The same information was also posted

to the distribution lists of four stroke patient and caregiver

groups. Details of the project were presented at 20 stroke groups/

clubs within the 11 mainland NHS areas in Scotland, distributed

at group meetings within the three island NHS areas in Scotland

and presented at eight stroke-related health professional meetings

or conferences (including the Scottish Stroke Nurses Forum and

the UK Stroke Forum).

We gathered 548 treatment uncertainties. These came from 15

stroke groups/clubs (183/548), 22 individual stroke survivors

(77/548), four individual carers (21/548), four health professional

groups/meetings (37/548), 61 individual health professionals

(198/548), and three guidelines/research recommendations (32/

548).

Check treatment uncertaintiesThe 548 submitted questions were checked: 162 were excluded as

they were not questions about the effect of treatment; the remain-

ing 386 were merged into 228 ‘indicative’ questions. Two were

removed as they were considered to have been answered by

research; these questions were related to the effectiveness of stroke

unit care and low molecular weight heparin for prevention of

deep venous thrombosis. Thus, there were 226 unique unan-

swered treatment uncertainties.

Interim prioritizationNinety-seven respondents (42 stroke survivors/caregivers and 55

health professionals) selected their 10 personal highest priorities

from the list of 226. We found substantial disagreement between

the ‘top’ questions of stroke survivors/caregivers and health pro-

fessionals. We identified the top questions of stroke survivors/

caregivers and the top questions from health professionals (we did

this based on both number of votes and assigned scores). We took

the top 12 questions from: (i) stroke survivors/caregivers number

of votes; (ii) stroke survivors/caregivers assigned scores; (iii)

health professional number of votes; and (iv) health professional

assigned scores. We merged all these questions, resulting in 24



4 Vol ••, •• 2012, ••–••

unique questions, representing the 24 shared treatment uncer-

tainties. These 24 treatment uncertainties are listed in Table 1.

Prior to the final priority setting stage, each uncertainty was allo-

cated a unique identification code (ID), from ID-A to ID-X: these

codes are used to refer to individual uncertainties within the text.

Final priority settingTwenty-eight people (16 stroke survivors/caregivers and 12

health professionals) attended a final consensus meeting on 16

November 2011. Health professionals included stroke physician

(1), nurse (1), physiotherapist (2), occupational therapist (2),

speech and language therapist (2), orthotist (1), orthoptist (1),

neuropsychologist (1), and social worker (1). Participants worked

in three independent groups to agree a group priority ranking of

the 24 shared treatment uncertainties. The ranked order from

each of the three groups was summed to give a total ‘score’, and a

‘combined ranking’ then determined based on the rank order of

this score. The ranked priorities from the three independent

groups and combined scores and ranks are displayed in Table 1.

There was a high level of agreement between the three groups.

Eight of the questions placed in the top 10 of Group A were in the

combined top 10; this was 7 for Group B and 9 for Group C. There

were only four individual questions for which one or more indi-

vidual group had ranked the question in their top 10, but the

combined ranking was greater than 10 (IDs-K, O, R, and W).

These questions were each discussed by the group until consensus

was reached on an agreed ranking for the top 10 questions.

The question with combined group ranking 7 (ID-V) was

debated at length. Consensus was reached that, while this ques-

tion was of very high priority to anyone experiencing a stroke,

it was beyond the defined scope of this PSP and – as such – ought

not to be placed in the top 10. It was agreed to move this question

to an effective ranked position of 11.

There was discussion around the generality of many of the

questions and the relationship with questions that had a more

specific focus. There was the explicit and conscious decision by

the consensus meeting participants that if there were more than

one question around the same problem or impairment after

stroke, then the more general question ought to be represented in

the top 10. These decisions are documented in Table 1.

It was agreed to insert a definition into the text of one question

and to add a list of potential interventions to a further three

questions. These were explicit measures taken to represent uncer-

tainties about specifically named treatments that did not reach the

top 10 within some of the more general top 10 questions. The

amended text of the questions is noted within Table 1.

By the close of the consensus meeting, there was unanimous

consensus on the top 10 shared research priorities.

Discussion

The top 10 research priorities relating to life after stroke have

been identified using a rigorous and person-centered approach.

This top 10 has been generated with equitable participation

from stroke survivors, caregivers, and health professionals and

therefore reflects what matters most to people affected by stroke.

Six of the top 10 research priorities are focused on specific

stroke-related impairments, including cognition, aphasia, vision,

upper limb, mobility, and fatigue, which would be classified as

‘body functions’ within the International Classification of Func-

tion, Disability and Health (ICF) framework (23). Three of the

top 10 are related to more social aspects of ‘living with stroke’

including one on coming to terms with long-term consequences

(ID-Q), one relating to confidence (ID-P), and one on helping

stroke survivors and their families ‘cope’ with speech problems

(ID-D). These three topics would be classified as ‘activities and

participation’ within the ICF framework. The remaining research

priority is related to the secondary consequences of stroke and

subsequent stroke prevention (the benefits of exercise and fitness,

ID-W), which could include aspects of both body functions and

environmental factors within the ICF framework.

During the consensus meeting, there was specific discussion

about the inclusion of the three questions within the activities and

participation category. For example, there was particular debate

around the question about improving confidence (ID-P); health

professionals suggested that this could perhaps be taken out of the

top 10 (as the combined group ranking placed it at 9=), but stroke

survivors reasoned that there was little point providing rehabili-

tation aimed at impairments if the patient did not have the con-

fidence to use the skills being learned during rehabilitation and

participate in daily activities. The example given was that speech

and language therapy to improve communication was ‘pointless’

if the person lacked the confidence to go into a shop and ask for

a newspaper. This priority setting process clearly highlighted that

research into social aspects of life after stroke is of equal priority

to stroke survivors as research into stroke-related impairments

and ought to be included on the international stroke research

agenda.

Scope of these prioritiesThe focus of this PSP was ‘life after stroke’. However, no limitation

was placed on the types of questions that could be submitted to

this process, nor to the time in the stroke pathway these questions

should refer to. As a consequence, a number of questions were

submitted and included that related to emergency, prevention,

and critical care medical research. A number of questions on these

topics were within the final shared Top 24, and one question

about the avoidance of delayed diagnosis (ID-V) was placed in the

top 10 by two of the independent groups at the consensus con-

ference. After significant discussion, the whole group reached

agreement that this question was beyond the defined scope of

‘life after stroke’ and should not be in the final top 10. During this

discussion, consensus was reached that, while research into emer-

gency, prevention, and critical care medical research extremely

important, the scope of the PSP had been defined as ‘life after

stroke’, and these top 10 priorities should directly relate to this

research agenda and not to the emergency, prevention, or critical

care research agendas.

Generality of prioritiesDuring the process of merging similar questions, a number of

subjective decisions had to be made. For many topics, there were

some broad ‘overarching’ questions (e.g. ID-G, what are the best



Vol ••, •• 2012, ••–•• 5

Tabl

e1

Con

sens

uson

shar

edto

p10

.Ind

ivid

ualg

roup

prio

ritiz

atio

n,co

mbi

ned

grou

psc

ore

and

rank

ing,

and

key

note

sre

latin

gto

shar

edto

p24

trea

tmen

tunc

erta

intie

s,as

dete

rmin

edat

Con

sens

usM

eetin

g

IDTr

eatm

ent

unce

rtai

nty

Gro

upA

rank

ing

Gro

upB

rank

ing

Gro

upC

rank

ing

Com

bine

dgr

oup

scor

e

Com

bine

dgr

oup

rank

ing

Fina

lagr

eed

top

10(a

ndke

yno

tes

from

cons

ensu

sm

eetin

g)

MW

hat

are

the

best

way

sto

impr

ove

unde

rsta

ndin

g(c

ogni

tion)

afte

rst

roke

a ?3

33

91

1Q

Wha

tar

eth

ebe

stw

ays

ofhe

lpin

gpe

ople

com

eto

term

sw

ithth

elo

ng-t

erm

cons

eque

nces

ofst

roke

?1

18

102

2

AW

hat

are

the

best

way

sto

help

peop

lere

cove

rfr

omap

hasi

a?10

51

163

3G

Wha

tar

eth

ebe

sttr

eatm

ents

for

arm

reco

very

and

func

tionb ?

77

519

44

SW

hat

are

the

best

way

sto

trea

tvi

sual

prob

lem

saf

ter

stro

ke?

810

220

55

XW

hat

are

the

best

way

sto

man

age

and/

orpr

even

tfa

tigue

?13

26

216

6I

Wha

tar

eth

ebe

sttr

eatm

ents

toim

prov

eba

lanc

e,ga

it,an

dm

obili

tyc ?

519

428

87

DH

owca

nst

roke

surv

ivor

san

dfa

mili

esbe

help

edto

cope

with

spee

chpr

oble

ms?

615

1031

9=8

PW

hat

isth

ebe

stw

ayto

impr

ove

confi

denc

eaf

ter

stro

ked ?

144

1331

9=9

WA

reex

erci

sean

dfit

ness

prog

ram

sbe

nefic

iala

tim

prov

ing

func

tion

and

qual

ityof

life

and

avoi

ding

subs

eque

ntst

roke

?23

818

4917

=10

VW

hat

isth

ebe

stw

ayto

avoi

dde

laye

ddi

agno

sis

ofst

roke

?2

137

227

[11]

EW

hat

isth

ebe

stph

ysio

ther

apy

regi

me

for

reco

very

afte

rst

roke

?11

614

319=

(inco

rpor

ated

into

num

ber

7)R

Are

stro

kegr

oups

/clu

bshe

lpfu

lat

impr

ovin

gth

ese

nse

ofw

ell-b

eing

amon

gst

roke

surv

ivor

s?4

920

3312

(inco

rpor

ated

into

num

ber

9)

OH

owca

nm

emor

ypr

oble

ms

afte

rst

roke

beim

prov

ed?

1511

935

13(in

corp

orat

edin

tonu

mbe

r1)

LW

hat

are

the

best

trea

tmen

tsfo

rsp

astic

ity?

1714

1546

14(c

onsi

dere

dto

beco

vere

dby

4an

d7)

BW

hich

spee

chan

dla

ngua

geth

erap

ytr

eatm

ents

wor

kbe

stfo

rap

hasi

a?19

1217

4815

=(c

onsi

dere

dto

beco

vere

dby

3)N

Wha

tis

the

best

way

toad

dres

s‘f

alse

belie

fs’

(impa

ired

perc

eptio

n)af

ter

stro

ke?

2116

1148

15=

(con

side

red

tobe

cove

red

by1)

UW

hich

trea

tmen

tsar

eth

ebe

stat

prev

entin

gst

roke

and

subs

eque

ntst

roke

?12

1819

4917

=C

Wha

tis

the

best

amou

ntan

din

tens

ityof

spee

chan

dla

ngua

geth

erap

yfo

rap

hasi

a?18

1716

5119

(con

side

red

tobe

cove

red

by3)

KLe

gsp

lints

toim

prov

eba

lanc

ean

dw

alki

ngaf

ter

stro

ke?

922

2152

20=

(inco

rpor

ated

into

num

ber

7)T

Whi

chor

thop

tictr

eatm

ents

are

best

atim

prov

ing

qual

ityof

life

afte

rst

roke

?20

2012

5220

=(c

onsi

dere

dto

beco

vere

dby

5)H

Upp

erlim

bsp

lints

impr

ove

arm

func

tion

and

prev

ent

com

plic

atio

nsaf

ter

stro

ke?

1624

2363

22(in

corp

orat

edin

tonu

mbe

r4)

FBo

tox

toim

prov

efu

nctio

nin

activ

ities

ofda

ilyliv

ing?

2221

2467

23(in

corp

orat

edin

tonu

mbe

r4)

JA

udito

ryfe

edba

ckto

impr

ove

wal

king

,in

depe

nden

ce,

and

qual

ityof

life

afte

rst

roke

?24

2322

6924

(inco

rpor

ated

into

num

ber

7)

a Defi

nitio

nad

ded

sofin

alun

cert

aint

yte

xtis

‘Wha

tare

the

best

way

sto

impr

ove

cogn

ition

afte

rstr

oke?

(Cog

nitio

n:th

efu

nctio

nof

proc

essi

ngin

form

atio

nan

dap

plyi

ngkn

owle

dge.

Func

tions

incl

ude

proc

esse

sre

quiri

ngth

ough

tan

din

telli

genc

e,su

chas

atte

ntio

n,pe

rcep

tion,

lear

ning

,m

emor

y,co

mpr

ehen

sion

,ju

dgm

ent

and

deci

sion

mak

ing)

’.b P

oten

tial

inte

rven

tions

liste

dso

final

unce

rtai

nty

text

is‘W

hat

are

the

best

trea

tmen

tsfo

rar

mre

cove

ryan

dfu

nctio

n?(in

clud

ing

visu

alfe

edba

ck,

virt

ual

real

ity,

bila

tera

ltr

aini

ng,

repe

titiv

eta

sktr

aini

ng,

imag

ery/

men

talp

ract

ice,

splin

ting,

elec

trom

echa

nica

land

robo

t-as

sist

edar

mtr

aini

ngan

dbo

tulin

umto

xin)

’.c P

oten

tiali

nter

vent

ions

liste

dso

final

unce

rtai

nty

is‘w

hat

are

the

best

trea

tmen

tsto

impr

ove

bala

nce,

gait

and

mob

ility

?(in

clud

ing

phys

ioth

erap

y,ga

itre

habi

litat

ion,

visu

alan

dau

dito

ryfe

edba

ck,e

lect

rical

stim

ulat

ion,

diff

eren

tty

pes

ofan

kle

foot

orth

oses

and

elec

trom

echa

nica

lass

iste

dga

ittr

aini

ng)’.

d Pot

entia

lint

erve

ntio

nslis

ted

sofin

alun

cert

aint

yte

xtis

‘Wha

tar

eth

ebe

stw

ays

toim

prov

eco

nfide

nce

afte

rst

roke

?(in

clud

ing

stro

kecl

ubs/

grou

ps,

offe

ring

supp

ort,

one-

to-o

nein

put

and

re-s

killi

ng)’.



6 Vol ••, •• 2012, ••–••

treatments for arm recovery and function?) but also some specific

questions often focused on a specific intervention (e.g. ID-H,

upper limb splints to improve arm function and prevent compli-

cations after stroke?). The decision of the steering group during

this phase was that it was not appropriate to merge the more

specific questions within the broader questions, as these specific

questions were important to consider and may be of a higher

priority to an individual stroke survivor, caregiver, or health pro-

fessional than a more general question. However, during the con-

sensus meeting, the participants made the explicit and conscious

decision that if there were more than one question around the

same problem or impairment after stroke, then the more general

question ought to be represented in the top 10. This was a prag-

matic approach taken by the consensus meeting participants in

order to reflect a wider range of stroke problems and impairments

within the top 10.

As a consequence of the decision to have the more general

questions in the top 10, most – if not all – of the top 10 questions

will need clarification and definition prior to research being insti-

gated. Thus, the top 10 questions are more reflective of the top

research ‘topics’ than of research questions per se. Nevertheless,

these identified research topics are those which are of greatest

importance to stroke survivors, caregivers, and health profession-

als and – as such – we believe it is appropriate for funders and

researchers to prioritise research around these topics.

Research relating to the top 10 questions could legitimately

include clinical and basic investigations by many disciplines and

may include basic neurology and pathology research as well as

applied rehabilitation research. However, it is important that

researchers do not lose sight of the fact that these questions reflect

what is of greatest importance to stroke survivors, caregivers, and

health professionals, and all proposals should consider the direct

relevance to the clinical and rehabilitation settings.

Strengths and weaknesses of this prioritysetting processFailure to limit the submitted treatment uncertainties to any

one point in the stroke pathway (e.g. rehabilitation) meant that a

number of submitted treatment uncertainties were effectively

beyond the defined scope of this PSP. However, these submitted

questions were not removed from the process. Maintaining in the

process questions that were beyond the scope of this project may

arguably have ‘diluted’ the responses to this process. However,

we argue that our process was strengthened by the fact that we

enabled participants to use their own interpretation of what ‘life

after stroke’ included. We feel that it was more appropriate for

the consensus meeting participants to have debated and reached

consensus on this issue, rather than for the steering group to have

made arbitrary decisions on a case-by-case basis.

A key strength of our work was the rigor of the processes we

implemented to facilitate equitable involvement of stroke survi-

vors, caregivers, and health professionals. The applicability and

relevance of the priorities arising from a process such as this are

dependent on the submission of representative and comprehen-

sive research questions and the equitable involvement of stroke

survivors, caregivers, and health professionals at all stages of the

process. We achieved approximately equal numbers of responses

from stroke survivors and health professionals at all stages of the

process. There were differences in the mode of submission of

stroke survivors and health professionals, with more stroke sur-

vivor responses being collected at face-to-face stroke groups and

clubs, and more health professional responses being returned by

e-mail, demonstrating the added benefit of our face-to-face out-

reach visits to stroke groups and clubs. Despite our attempts to

involve caregivers, we had relatively poor responses from care-

givers and acknowledge that further work is required to gain

equitable involvement from this group.

Conclusions

The top 10 research priorities relating to life after stroke have been

identified using a rigorous and person-centered approach. These

present clear directions for future research relating to life after

stroke and should be valuable for informing the funding of future

stroke research.

Acknowledgements

This project was completed as part of the DORIS project, which

was supported by funding provided by the Scottish Government’s

National Advisory Committee for Stroke. This project was carried

out by the Nursing Midwifery and Allied Health Professions

(NMAHP) Research Unit, which is supported by the Scottish

Government Health Directorate’s Chief Scientist Office. The work

presented here represents the view of the authors and not neces-

sarily those of the funding bodies.

This project was completed in partnership with Chest Heart

and Stroke Scotland and The Stroke Association in Scotland.

References1 INVOLVE. Supporting public involvement in NHS, public health and

social care research. Hampshire: UK; 2011. Available at http://www.invo.org.uk/ (accessed 23 February 2011).

2 Staley K. Exploring Impact: public involvement in NHS, public healthand social care research. INVOLVE, Eastleigh. 2009.

3 Chalmers I, Glaszio P. Avoidable waste in the production and report-ing of research evidence. Lancet 2009; 374:86–9.

4 Partridge N, Scadding J. The James Lind Alliance: patients and clini-cians should jointly identify their priorities for future trials. Lancet2004; 364:1923–4.

5 Oliver SR. How can health service users contribute to the NHSresearch and development programme? BMJ 1995; 310:1318–20.

6 Elwyn G, Crowe S, Fenton M et al. Identifying and prioritising uncer-tainties: patient and clinician engagement in the identification ofresearch questions. J Eval Clin Pract 2010; 5:1–5.

7 The Stroke Association. The Stroke Association Corporate Strategy2010 to 2015. March 2010.

8 Boote T, Dalgleish M, Freeman J, Jones Z, Miles M, Rodgers H. ‘Butis it a question worth asking?’ A case study describing how publicinvolvement can lead to researchers’ ideas being abandoned. Int JStroke 2011; 6(Suppl. 2):58.

9 Partridge N, Scadding J. The James Lind Alliance: patients and clini-cians should jointly identify their priorities for clinical trials. Lancet2004; 364:1923–4.

10 Opar A. Lind guidelines offer a checklist for research priorities. NatMed 2010; 16:620.



Vol ••, •• 2012, ••–•• 7

11 Petit-Zeman S, Firkins L, Scadding JW. The James Lind Alliance:tackling research mismatches. Lancet 2010; 201:667–9.

12 Cowan K, Oliver S. The James Lind Alliance Guidebook. Oxford: UK;2012 Available at http://www.jlaguidebook.org/ (accessed 15 March2012).

13 Lophatananon A, Tyndale-Biscoe S, Malcolm E et al. The James LindAlliance approach to priority setting for prostate cancer research: anintegrative methodology based on patient and clinician participation.Br J Urol Int 2011; 108:1040–3.

14 Eleftheriadou V, Whitton M, Gawkrodger D et al. Future research intothe treatment of vitiligo: where should our priorities lie? Resultsof the vitiligo priority setting partnership. Br J Dermatol 2011; 164:530–6.

15 McClurg D. The James Lind Alliance and urinary incontinence. J AssocChart Physiotherapists Womens Health 2010; Spring:106.

16 Lloyd K, Rose D, Fenton M. Identifying uncertainties about theeffects of treatments for schizophrenia. J Ment Health 2006; 15:263–8.

17 Jones J, Hunter D. Consensus methods for medical and health servicesresearch. BMJ 1995; 311:376–80.

18 Cowan K. The James Lind Alliance – tackling treatment uncertaintiestogether. J Ambul Care Manag 2010; 33:241–8.

19 DORIS. Database of Research in Stroke. Glasgow: UK; 2012. Availableat http://www.askDORIS.org (accessed 8 May 2012).

20 Richardson WS, Wilson MC, Nishikawa J, Hayward RSA. The well-built clinical question: a key to evidence-based decisions. ACP J Club1995; 123:A12–3.

21 Brown P, Brunnhuber K, Chalkidou K et al. How to formulateresearch recommendations. BMJ 2006; 333:804–6.

22 SIGN 118. Management of patients with stroke: rehabilitation,prevention, management of complications, and discharge planning.A national clinical guideline. Edinburgh, UK: Scottish IntercollegiateGuidelines Network; 2010.

23 World Health Organization. International Classification of Function-ing, Disability and Health: ICF. Geneva, World Health Organization,2001.



8 Vol ••, •• 2012, ••–••

Documents

Top 10 research priorities relating to life after stroke - consensus from stroke survivors, caregivers, and health professionals