Cultural validation of the Addenbrooke’s Cognitive Examination Version III Urdu for the British Urdu-speaking population: a qualitative assessment using cognitive interviewing

Abstract

Objectives
Our research determined whether the Addenbrooke’s Cognitive Examination Version III (ACE-III) Urdu eliminated cultural bias through a qualitative assessment of its understanding and acceptability within the British Urdu-speaking population, employing cognitive interviews.

Method
We aimed to recruit 25 participants fluent in speaking and writing Urdu, over the age of 60 years, able to give informed consent and who did not have a history of cognitive impairment. Participants were administered the ACE-III Urdu, and cognitive interviews were conducted, which involve obtaining verbal data on the individual’s perception of the assessment overall, their understanding of the mental processes behind how they interpreted questions within the assessment and how they produced appropriate responses. This allows us to gauge the participants’ overall thoughts on the Urdu ACE-III before applying question-formatted prompts to every ACE-III Urdu item.

Results
We recruited 25 participants, 12 women (48%), ranging from ages 60 years to 85 years (M=69.12, SD=6.57), all from Greater Manchester. Participants came from varied socioeconomic backgrounds, with 22 identifying as Pakistani, one as British Pakistani and two as East African. Across 19 ACE-III Urdu items, 7 required changes based on participant feedback: item 5a: fluency; items 6, 18 and 19: memory; items 12 and 13: language; and item 17: visuospatial abilities.
The need for some of these changes was realised after 21 participants, due to persistently reoccurring issues, and these were applied before the last four participants. Overall, the ACE-III Urdu was considered easy and straightforward by all 25 participants, who understood items and felt the ACE-III Urdu was appropriate, not just for them, but for British Urdu speakers in general.

Conclusion
Our cognitive interviews determined the ACE-III Urdu was acceptable, especially with regards to cultural context, but further changes were made to ensure understanding. Therefore, we adapted the ACE-III Urdu in accordance with feedback, resulting in our finalised version being culturally validated.

Full text

1
MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
Cultural validation of the
Addenbrooke’s Cognitive Examination
Version III Urdu for the British Urdu-
speaking population: a qualitative
assessment using cognitive interviewing
Nadine Mirza, Maria Panagioti, Waquas Waheed
To cite: MirzaN, PanagiotiM,
WaheedW. Cultural validation
of the Addenbrooke’s Cognitive
Examination Version III Urdu
for the British Urdu-speaking
population: a qualitative
assessment using cognitive
interviewing. BMJ Open
2018;8:e021057. doi:10.1136/
bmjopen-2017-021057
► Prepublication history and
additional material for this
paper are available online. To
view these les, please visit
the journal online (http:// dx. doi.
org/ 10. 1136/ bmjopen- 2017-
021057).
Received 7 December 2017
Revised 25 September 2018
Accepted 2 November 2018
Centre for Primary Care, The
University of Manchester,
Manchester, UK
Correspondence to
DrWaquas Waheed;
Waquas. waheed@ manchester.
ac. uk
Research
© Author(s) (or their
employer(s)) 2018. Re-use
permitted under CC BY-NC. No
commercial re-use. See rights
and permissions. Published by
BMJ.
ABSTRACT
Objectives Our research determined whether the
Addenbrooke’s Cognitive Examination Version III (ACE-
III) Urdu eliminated cultural bias through a qualitative
assessment of its understanding and acceptability within
the British Urdu-speaking population, employing cognitive
interviews.
Method We aimed to recruit 25 participants uent in
speaking and writing Urdu, over the age of 60 years, able
to give informed consent and who did not have a history
of cognitive impairment. Participants were administered
the ACE-III Urdu, and cognitive interviews were conducted,
which involve obtaining verbal data on the individual’s
perception of the assessment overall, their understanding
of the mental processes behind how they interpreted
questions within the assessment and how they produced
appropriate responses. This allows us to gauge the
participants’ overall thoughts on the Urdu ACE-III before
applying question-formatted prompts to every ACE-III Urdu
item.
Results We recruited 25 participants, 12 women
(48%), ranging from ages 60 years to 85 years
(M=69.12, SD=6.57), all from Greater Manchester.
Participants came from varied socioeconomic
backgrounds, with 22 identifying as Pakistani, one
as British Pakistani and two as East African. Across
19 ACE-III Urdu items, 7 required changes based on
participant feedback: item 5a: fluency; items 6, 18
and 19: memory; items 12 and 13: language; and item
17: visuospatial abilities. The need for some of these
changes was realised after 21 participants, due to
persistently reoccurring issues, and these were applied
before the last four participants. Overall, the ACE-III
Urdu was considered easy and straightforward by all 25
participants, who understood items and felt the ACE-III
Urdu was appropriate, not just for them, but for British
Urdu speakers in general.
Conclusion Our cognitive interviews determined the
ACE-III Urdu was acceptable, especially with regards
to cultural context, but further changes were made to
ensure understanding. Therefore, we adapted the ACE-
III Urdu in accordance with feedback, resulting in our
finalised version being culturally validated.
Dementia affects 46.8 million people glob-
ally, with 5%–7% prevalence in those over
60 years.1 2 In the UK, this results in over
800 000 suffering from dementia, predicted
to rise to 1 million in 2025,3 becoming one of
the top five causes of death.4
Ethnic minority groups are at high risk
for dementia with the increasing number
of elderly in these groups5 and a higher
prevalence among those who do not speak
English.6 South Asians, accounting for 6.3%
of the UK population,7 are no exception, with
approximately 17.7% of British South Asians
over the age of 50 years.8 Estimations of this
population and rates of dementia show that
South Asian elderly between 65 years and
Strengths and limitations of this study
►We have demonstrated a robust cultural validation
technique that is not time consuming and is easy to
conduct with no specialist expertise.
►Our research was conducted in a culturally sensitive
manner.
►We recruited participants from various educational
backgrounds and also had representation from both
genders, and the number of participants we recruit-
ed ensured thematic saturation.
►During recruitment, a more robust screening pro-
cess to assess the participants’ cognitive health
would have been preferable but were limited due to
a paucity of psychometrically validated cognitive as-
sessment scales available for British Urdu speakers
and would have run the risk of false positive and
negative scores. Our convenience sample could also
not determine that our sample was statistically rep-
resentative of our target.
►Though we were able to identify a pattern of dif-
culty in ve items after 21 interviews and address
these, this should ideally have been recognised
earlier and perhaps applied half way through the
sample instead.
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

2MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
79 years have the highest prevalence for dementia in the
UK and second highest for those above 80 years.9
This prominence is attributed to relevant health issues
common in British South Asians, such as vascular diseases
and type 2 diabetes,10 but we also see a lack of suitable
cognitive assessments that interfere with the detection
and intervention of dementia.11 Diagnoses via cognitive
assessments lead to higher rates of false positive scores
and overestimated cognitive impairment12 within British
South Asians,13 and of false negative scores, so assess-
ments fail to detect dementia until it is too late for early
interventions.11 14
One reason is that cognitive assessments are designed
in Western countries for specific English-speaking popu-
lations,15 standardised on memory disorder clinic outpa-
tients, the vast majority are well-educated Caucasians.16
Therefore, they are ill suited to the needs of British
South Asian elderly who are accustomed to cultures of
their native countries, with 73% relying on their first
language.17
Culture impacts performance on cognitive assessments,
influencing perception of and responses to assessment
items,18 and translation does not account for this.13 As
British South Asians are scoring lower on assessments
that have different cultural contexts than the ones they
are familiar with, there is a need for assessments for this
population, countering effects of cultural bias that arise
when a target population differs from the population the
assessment was originally designed for.13 18
We translated and culturally adapted a cognitive
assessment for the British Urdu-speaking population,
The Addenbrooke’s Cognitive Examination Version III
(ACE-III), which is a gold standard diagnostic tool for
detecting dementia19 20 (see table 1). It assesses five
Table 1 Items of the Addenbrooke’s Cognitive Examination Version III
Item number Task/question
1: Attention – orientation Ask the day, date, month, year, season, oor, street/hospital, town, county and country.
2: Attention – registration Say the words lemon, key and ball and ask them to repeat and try to remember.
3. Attention – concentration Ask to take 7 away from 100 and keep taking 7 away from the new number for 5 trials (Serial 7’s).
4. Memory –recall Ask for the three words from 2. Attention – registration.
5a. Fluency –letters Ask for as many words as they can think of starting with the letter ‘P’, not including names of pronouns, in 1 min.
5b. Fluency –animals Ask for the names of as many animals as they can think of in 1 min.
6. Memory –anterograde Say the name and address ‘Harry Barnes, 73, Orchard Close, Kingsbridge, Devon’ and ask them to repeat and try
to remember.
7. Memory –retrograde Ask for the name of the current Prime Minister, name of the woman who was Prime Minister, name of the USA
president and name of the USA president who was assassinated in the 1960s.
8. Language –comprehension Place a pencil and paper in front. Ask to ‘place the paper on top of the pencil’, ‘pick up the pencil but not the
paper’ and ‘pass me the pencil after touching the paper’.
9. Language –writing Ask to write two or more complete sentences about their last holiday/weekend/Christmas, without using
abbreviations.
10. Language –repetition Say the words caterpillar, eccentricity, unintelligible and statistician and ask them to repeat.
11. Language –repetition Say the proverbs ‘All that glitters is not gold’ and ‘A stitch in time saves nine’ and ask them to repeat.
12. Language –naming Show 12 images and ask them to name each.
13. Language –comprehension Ask to point to ‘the one which is associated with the monarchy’, ‘the one which is a marsupial’, ‘the one which is
found in the Antarctic’ and ‘the one which has a nautical connection’ from the 12 images provided.
14. Language –reading Ask them to read the words sew, pint, soot, dough and height.
15a. Visuospatial abilities –innity diagram Ask them to copy the following:
15b. Visuospatial abilities –wire cube Ask them to copy the following:
15c. Visuospatial abilities – clock Ask them to draw a clock face with numbers and the hands at ten past ve.
16. Visuospatial abilities Ask them to count the number of dots without pointing.
17. Visuospatial abilities Ask them to identify the fragmented letters K, M, A and T.
18. Memory –recall Ask for the three words from 6. Memory – anterograde.
19. Memory – recognition For each word of the name and address that could not be recalled, give the options listed and ask to identify which
word it was.
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

3
MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
cognitive domains: attention, memory, fluency, language
and visuospatial abilities.20 Through methodology
described elsewhere, we translated the ACE-III into Urdu
(Nadine Mirza, the translation and cultural adaptation
of the ACE-III for a British Urdu-speaking population),
a popular South Asian language that is the fourth most
spoken in the UK, and culturally adapted it for British
Urdu speakers.7
This paper details the research, following the
above-mentioned translation and cultural adaptation
process, which was required to determine whether
the ACE-III Urdu eliminated cultural bias through a
cultural validation process (Nadine Mirza, The trans-
lation and cultural adaptation of the ACE-III for a
British Urdu-speaking population). Cultural validation
differs from psychometric validation in that it does not
determine how the assessment compares against other
standardised assessments or within a healthy popula-
tion versus a clinical one and demonstrates whether
an assessment is detecting or measuring what it was
designed to do so.19
Instead, cultural validation determines whether the
translation is understandable and whether the cultural
adaptation allows for understanding and acceptability of
the assessment within a target population (Nadine Mirza,
The translation and cultural adaptation of the ACE-III for
a British Urdu-speaking population). To ensure robust
translated and culturally adapted assessments, they should
undergo a cultural validation followed by a psychometric
validation.
This research conducted cultural validation through a
qualitative assessment of the understanding and accept-
ability of the ACE-III Urdu within the target population
by employing cognitive interviews. The cognitive inter-
viewing process was originally developed to identify and
amend errors in survey questions and is now used by
survey centres, government agencies, research firms and
even lay persons to develop assessments.21 22 It determines
whether an assessment is generating the responses it was
designed for.21
The process involves administering the assessment and
obtaining verbal data on the individual’s perception of
the assessment overall, their understanding of the mental
processes behind how they interpreted questions within
the assessment and how they produced appropriate
responses.21 The data can be obtained through a think
aloud approach, which involves minimal intervention in
facilitating the individuals’ responses, or through verbal
prompting, which uses direct questions.21 23
Our review of translations and cultural adaptations
of the ACE-III24 found no adaptors undertaking such
a detailed cultural validation process. Therefore, we
assessed whether our ACE-III Urdu was suitable and
acceptable within the British Urdu-speaking population
through administering the ACE-III Urdu and conducting
cognitive interviews with lay persons from our target
population.
METHODS
Participants
This research was to be conducted within British South
Asian communities of Greater Manchester. We aimed to
recruit 25 participants25 fluent in speaking and writing
Urdu, over the age of 60 years, able to give informed
consent and who did not have a history of cognitive
impairment.
The principal researcher, NM, was matched to the target
population, fluent in English and Urdu and familiar with
British South Asian culture. She underwent training for
recruiting ethnic minorities to research in a culturally
sensitive manner.26
Participants would be voluntarily recruited via conve-
nience sampling from the local Pakistani Community Day
Centre, community libraries, mosques, shopping centres
and through snowball sampling. Participants would be
provided with an information sheet, available in Urdu and
English, and in accordance with University of Manchester
Research Ethics Committee policy, and the cultural sensi-
tivity training26 would be given 24 hours to decide if they
wished to participate. NM would contact them to confirm
a location, date and time.
Materials
Participants would be provided with consent forms and
demographics sheets, available in Urdu and English. A
pen is needed for scoring, and a timer, pencil and paper
are required for the assessment.
The ACE-III Urdu was developed as part of our
Dementia in Ethnic Minorities study for a British Urdu-
speaking population (Nadine Mirza, The translation and
cultural adaptation of the ACE-III for a British Urdu-
speaking population). This was through a robust method-
ology that required developing guidelines on translating
and culturally adapting the ACE-III, with rationale justi-
fying changes, and then using these guidelines along with
qualitative data from focus groups with Urdu-speaking
lay persons and experts within the field to translate and
culturally adapt the ACE-III for British Urdu speakers
(Nadine Mirza, The translation and cultural adaptation
of the ACE-III for a British Urdu-speaking population).
For our cognitive interviews, we planned to adopt
a verbal prompting approach, which is deemed less
burdensome for participants.21 This allows for guiding
participants through thought processes they underwent
while attempting the assessment. We aimed to use verbal
probes and prompts about comprehension and meaning
and requests to paraphrase questions.21 23 27 This would
be facilitated by a semistructured interviewing schedule
translated into Urdu (see supplementary file 1 appendix
1).
Questions gauge the participants’ overall thoughts on
the Urdu ACE-III before applying question-formatted
prompts to every ACE-III Urdu item as follows:
1. What they thought the item had asked and why.
2. If they can paraphrase the item.
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

4MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
3. What mental processes were behind their responses
and why.
4. How difficult the item was.
5. Any other suggestions regarding its translation and
cultural adaptation.
Procedure and analysis
NM aimed to conduct interviews at the Pakistani Commu-
nity Centre or the participants’ residences. After informed
consent is obtained, participants would be administered
the ACE-III Urdu over a period of 20–30 min. Participants
would be provided with their overall score and a break-
down of their scores across cognitive domains.
Following this, NM would undertake a retrospective
probing technique; participants would be administered
the cognitive interview after they attempted the ACE-III
Urdu, allowing them to attempt the assessment in its
natural flow without interference from probing.21 On
completion, participants would be asked if they had
questions.
Simultaneously, if reoccurring issues with certain
ACE-III Urdu items emerged across participants at this
data collection stage, changes according to participant
feedback would be applied to only those ACE-III Urdu
items, and they would be administered to the remaining
participants.
All interviews would be audio recorded and transcribed
verbatim, from Urdu to English where necessary, to allow
for familiarity with the data. A thematic analysis would
identify themes focusing on participant acceptability
and overall perceptions of the ACE-III Urdu. Transcripts
would be analysed on a semantic level, and data relevant
to ACE-III Urdu items were highlighted.28
Extracts would be regarded independently and grouped
such that every item had several extractions relating to it.
Groups of extracts would be reviewed to identify under-
lying themes relevant to participant understanding and
acceptability of the items of the ACE-III Urdu. Themes
would be reviewed and refined, with supporting quotes
from transcriptions.
Proposed suggestions for changes to the ACE-III Urdu
would be reviewed. Item by item it would be decided if
the culturally adapted version of an item needed to be
further adapted according to these suggestions, and
changes would be applied where needed.
Patient and public involvement
There would be no patient or public involvement.
RESULTS
We successfully recruited 25 participants, 12 female
(48%), from ages 60–85 years (M=69.12, SD=6.57), from
Greater Manchester (See table 2). Participants came from
varied socioeconomic backgrounds, with 22 identifying
as Pakistani, 1 as British Pakistani and 2 as East African.
Based on participant feedback, changes were applied to
finalise the ACE-III Urdu (see table 3).
Though the finalised version incorporated the feed-
back of all 25 participants’ cognitive interviews, after 21,
we became aware of reoccurring issues in responding
to items 6, 12, 13, 18 and 19. These were perceived as
Table 2 Participant characteristics
Pt First language
Second
language
Highest level of
education
1 Urdu Punjabiand
English
BSc
2 Punjabi Urdu BA
3 Urdu Punjabi GCSE
4 Punjabi English,
Urduand
Arabic
Graduate
5 Urdu English BA; LLB
6 Punjabi Englishand
Urdu
Matric
7 Urdu English BA; LLB
8 Punjabi English, Urdu,
Gujratiand
Swahili
FA (12 years)
9 Punjabi Urdu Matric (10 years)
10 Punjabi Englishand
Urdu
Sixth Level
11 Urdu Englishand
Punjabi
BA
12 Urdu Englishand
Punjabi
FA (12 years)
13 Urdu English Matric (10 years)
14 Punjabi Englishand
Urdu
Postgraduate
15 Urdu English BA
16 Punjabi Englishand
Urdu
Matric (10 years)
17 Punjabi English, Urdu,
Gujratiand
Swahili
GCSE
18 Punjabi Englishand
Urdu
CA
19 Punjabi Englishand
Urdu
BA
20 English Punjabiand
Urdu
MRCP
21 Punjabi Englishand
Urdu
Graduate
22 Punjabi Englishand
Urdu
Diploma
23 Punjabi Englishand
Urdu
Matric (10 years)
24 Urdu English MSc
25 Punjabi Englishand
Urdu
AAT
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

5
MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
consistently ambiguous or not understandable. We
decided to apply changes to these five ACE-III Urdu items
based on the participants’ feedback before resuming
cognitive interviews with the remaining four participants.
They reported better responses to these particular items,
finding them easy to understand, retain and repeat.
Overall perception
The ACE-III Urdu was considered an easy and straightfor-
ward assessment. Participants understood items and felt
the ACE-III Urdu was appropriate for them and British
Urdu speakers in general.
P12: I understood all of it. I understood everything
that you were saying. […] I found it easy. Not hard.
P24: It was easy. Straight forward, you know?
Everything was straightforward […] and there was
nothing vague about it.
P25: It’s okay. It’s just fine. Straight forward test.
Item 1: attention
This item was considered straightforward, asking for
simple information such as the date or the building the
individual is in.
P1: First task- that was just personal information […]
whether I know who I am, where I am. Whatever. Just
some basic information […] You couldn’t make it any
simpler.
When someone may not know answers to the questions,
participants proposed reasoning other than cognitive
impairment. One suggested remembering the date is
harder during the holidays as they are not attending to
weekly responsibilities, reducing importance of knowing
the exact date.
P19: During the holidays it’s a bit difficult. Otherwise,
during the weekdays the children go to school but
these days there are holidays.
Participants, especially women, were less aware of the
specifics of their address, such as the ‘county’, because
matters related to their address were handled by their
spouses.
P19: I’m telling you I never bothered. That’s why I- the
address, actually lots of things with me- I, you know
never paid attention I tell you. I don’t bother with
these things. I don’t pay attention because my hus-
band knows all these things. He tracks these things.
Table 3 Changes applied to the ACE-III Urdu
Item Suggestions Justification
5a: Fluency The letter کwas shown when giving
the example of the task.
Participants would become confused when listening to the demonstrative
example of the task, focusing on the letterک and confusing it with the similar
sounding letter ق. Showing it would allow participants to know which letter we
were referring to.
6: Memory The address was changed to
73, Station Road,
Stockport,
Cheshire.
The address, though a UK address, was deemed too difcult to pronounce for
the average Urdu-speaking person. Participants would struggle to repeat the
words, and this affected their ability to memorise them.
An address that used elements better known to the British Urdu-speaking elderly
was developed.
12: Language i. The image for a book was changed.
ii. The image for the suitcase was
changed.
iii. The image for the bear was
changed.
i. According to some participants, the image for a book could be mistaken for
a folder as the pages were not clear. This was changed to a clearer image of
a book.
ii. According to some participants, the image of the suitcase could be confused
with other objects such as a plain box or a toaster. This was changed to the
image of a more modern wheeled suitcase.
iii. The image of the bear was not clear and participants would confuse it for
other four-legged wild animals. This was changed to a more realistic and
clearer image of a bear.
13: Language Questions were rephrased slightly:
‘Which one is related to our head’and
‘These days, which one is related to
travel’.
The questions were rephrased slightly for clarication.
Participants would become confused when referring to ‘the head’ and were able
to understand better when changed to ‘our head’.
Participants would often indicate a camel when asked, ‘Which one is related to
travel’ due to historical context. Changing it to include ‘these days’, allowed for
participants to automatically consider the suitcase as the only option.
17: Visuospatial
Abilities
The letters were erased further. The letters were considered easily recognisable by participants at the current
level of erasure so more was required.
18: Memory Refer to Item 6: Memory. Refer to Item 6: Memory.
19: Memory Alternate address elements were
developed in line with the changes
made to the original address:
Station Place/Oak Road,
Kingsbridge/Longsight.
Refer to Item 6: Memory.
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

6MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
The education of the individual was also called into
question, as participants believed it could affect responses.
P5: This depends on whether they are educated, you
know? Yeah, those who know a little, then they will
give the answer quickly. Others will give it with a bit
of difficulty.
Item 2: attentionand item 4: memory
The objects ‘lemon’, ‘bell’ and ‘ball’ were regarded as easy
to repeat and remember, commonly known and average
Urdu speakers should be able to complete these items.
P11: They come in everyday use, in daily routine.
Children play with a ball. Average things come be-
fore you over and over again. Lemon is in the kitchen
every day. And the bell rings at the door every day.
P13: Lemons are used at home, the children play
with balls and bells are rung by visitors so these were
ordinary items.
P24: Well, they’re every day things. I mean, you always
say ‘lemon’. ‘Ball’ as you say, it’s a normal sort of word.
Item 3: attention
The ‘Serial 7’s task’ was considered slightly difficult by a
majority of the participants. Participants understood what
they were being asked to do and attributed difficultly to
the task itself.
P11: It’s not very difficult but obviously, in mathe-
matics, in counting […] add 7 and 3, that’s sort of
more easy and like this backwards? Yeah, I did have
to stress a little bit in the mind that I didn’t say the
wrong thing.
P13: I know this so it wasn’t hard for me but it will be
for people […], those who don’t know mathematics.
Those who perceived the task to be easy attributed it to
a past profession that involved numbers.
P24: Well, I’m an accountant so it was easy for me.
But for somebody else, you know they would start
thinking about it and count on their fingers. Do that
sort of thing.
One argument to suggest older Urdu speakers should
perform well on this task was the assumption that elderly
have more practice with mental calculations as they did
not use a calculator.
P19: It’s difficult but you have to concentrate because
you’re doing it off by heart. […]our previous gener-
ation did even better off by heart because back then
everyone memorised off by heart. The later genera-
tion can’t do things off by heart as long as we have
these laptops. Calculators they use.
Item 5a: uency
Participants felt the common belief was that Urdu
speakers, if fluent, should complete this item successfully.
P24: It’s not a difficult letter to- with the language you
know? Maybe someone who doesn’t speak Urdu so
much, you know, will find it difficult to struggle.
However, it was acknowledged that few words start with
(chay) چ as compared with other letters, presenting this
task as a challenge.
P3: چ is such a letter […] that very few words start
with. […] Well, I mean for the average person it’s a
bit difficult- a bit.
P13: I knew fewer words beginning with the letter چ
like ‘chai’ (tea, like ‘chup’ (be quiet).
P14: No it was just چ, which means it’s just unusual
after- not to say, yeah after a long time چ, I didn’t even
remember. Yes, it’s not that common (bay) ب (alaf)
ا،that is, yeah but چ is an unusual word.
Prior to the task, an example is given in the item using
the letter ک (kaaf). This example confused a few partici-
pants as it sounds very similar to ق (qaaf).
P11: Ugh Yes. With ‘kaaf’, understanding- there are
two ‘kaaf’. ‘Qaaf’ with the two dots and the stick ‘kaaf’
so for that- like if you show the picture to a patient,
that with this kaaf.
Item 5b: uency
This item was considered straightforward because we are
surrounded by various animals.
P3: Because animals, they’re coming and going.
P11: Did not find it difficult at all. Animals’ names
are often remembered by a person so.
P24: I just read out the ones that immediately come
to mind. Also, you come to a point, you know, you
start thinking, oh what kind of animal should? I say,
oh bird! You know reptiles, you remember thousands
of them, you know, or hundreds of animals you know.
You see them all the time on television or natural life.
Item 6: memory; item 18: memory; item 19: memory
There were contrasting opinions regarding elements
of these items, all of which required repeating, remem-
bering or recognising a name and UK address.
The name was regarded positively as ‘Haroon Butt’
is a common name that is acceptable and culturally
appropriate.
P5: Haroon Butt, this is a common name because
nowadays, what’s their name? The Butts rule in
Pakistan.
P24: It’s fairly common. Butt is- big community of
Butts here […] And Haroon is a common name.
Participants would make a connection to someone in
their life named Haroon to recall the name.
P11: The name Haroon for people I’ve heard often,
like in my family there is the name Haroon. In friends
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

7
MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
as well. So because of this reason maybe I did not find
the name difficult in remembering.
P19: Because this is my uncle’s name too, Haroon.
[…] It was a connection. Haroon Chaudhry is his
name. ‘Chaudhry Haroon Butt’ sahib is what I
thought.
Alternatively, the address was considered difficult to
pronounce, causing issues in repetition and poor recall
and recognition scores across the first 21 participants.
Participants stated an average Urdu speaker would not be
able to remember the address.
P5: Name is easy to remember but the address, 73
what is it? Orchard Close, that they might forget[…]
and secondly, Kings… Kingsbridge. They might for-
get that too.
At times it was different aspects that posed difficulties,
such as an area they may have never visited or heard of.
They felt these difficulties would extend to the general
British Urdu-speaking elderly.
P3: Orchard is rare- ugh a rare name. Like street
names you know? They’re easy names normally, more
often than not. Orchard Close is a very rare name.
P11: I forgot Devon, because Devon I hear very little.
Had it been Manchester I would not have forgotten.
[…] I feel that the average Urdu speaking people or
it’s possible that people slightly older in age than me
they- they might find it difficult with names like this.
P14: Kingsbridge. Never been there. If it had been
any other town, Leicester, Birmingham or to wher-
ever I’ve been- I’ve never been to this town you
see- bridge… so that’s why you see, normally I make
connections you see.
Participants put forward suggestions for more cultur-
ally appropriate addresses.
P5: To make it easier for example, have a Manchester
address. My suggestion would be to have a local ad-
dress, a main road. Or a more famous street for ex-
ample, Slade Lane, which everyone knows about. Or
Dickenson Road.
These suggestions were considered, and the address
was changed (see table 3). The updated version was
administered to the last four participants who found it
easy to pronounce, repeat and remember and deemed it
acceptable for other British Urdu speakers.
P24: Names are all common names. Station Road
is everywhere […] Every town, every village has a
Station Road.
Item 7: memory
The consensus was that success on this item depends on
one’s understanding of current affairs and knowledge,
which they should know while living in the UK, and lack
of this knowledge was not limited to British South Asians.
P5: David Cameron resigned and Theresa May came
along […] if you watch television or read the newspa-
pers those people would know.
P24: For me it wasn’t difficult at all obviously, knowing
these things. Again, its peoples’ knowledge. A lot
of English people on the street, you ask the Prime
Minister, you will find so many saying no idea. So,
especially this one because she’s so new.
‘Name of the current Prime Minister’
This was considered the most difficult question, but this
may be due to recent changes in the British government at
the time of this research (David Cameron stepping down
from the role of Prime Minister, replaced by Theresa
May). Participant’s felt that if enough time passed, this
question would become easier once the state of govern-
ment was more stable.
P3: No, no. Those are easy. It’s just- the UK has
changed so remembering May’s name is difficult.
After one or two years every person will remember.
P11: Did not find it difficult but because she just came
into power her name is not that familiar so because
of this.
‘Name of the rst female Prime Minister’
This was considered easier by participants due to the
historical significance of Margaret Thatcher, the first
female Prime Minister of the UK. Many participants
remembered the identifier ‘The Iron Lady’, which led to
the memory of the name Margaret Thatcher.
P1: Margaret Thatcher was the- the- a big thing. It
made big headlines, big sensation. People- majori-
ty hated her because she was Tory- because she was
posh.
‘Name of the USA president’
This was considered easy even if participants did not read
newspapers or watch the news on television, as the presi-
dent at this time, Barack Obama, had been in presidency
for two terms and is spoken about frequently.
P19: Because he- mention of America. Otherwise it’s
not a part of knowledge.
P22: Because it’s been 8 to 10 years with him.
‘Name of the princess who died in a car crash in the 1990s’
Participant’s considered this an easy question due to
Princess Diana’s prominence when she was alive and the
historical and global significance of her death.
P19: Diana is in history. When I was here earlier,
when I first got married I knew about her accident.
The history, that’s why.
P24: Diana, obviously you know she was popular.
People would know. Can’t find someone who has no
idea who she was.
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

8MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
Item 8: language
This item was deemed culturally appropriate with no
need for adaptation.
Item 9: language
Most participants focused on the topic of holidays and
wrote about Pakistan, where they would spend vacations.
P19: I wrote a very simple sort of obviously- you said
holidays so I wrote about holidays. So no it was not
very difficult.
Participants acknowledged this may not simple for all
of the British Urdu-speaking population as some of them
may not have the literacy skills needed.
P5: This depends, not everyone can do this task.
Someone who’s at least a little educated can do this.
Here, people my age, many are illiterate […] but oral-
ly they can say it but they don’t know how to write it.
Item 10: language
Though all participants completed the task there were
differing opinions regarding whether these words were
familiar enough to be generalised to the British Urdu-
speaking population or if they were complex and only
suitable for a specific group of Urdu speakers.
Some knew these words from childhood.
P11: I did not find it difficult because in my child-
hood my uncle used to ask me these. These are my
childhood memory.
P14: Because ugh childhood- I had memorised […]
a common word, so it’s easy you see. Yes, should do
easily- Urdu speaking should do. Not maybe people
who come from Punjab or haven’t been educated.
Others reflected on how a person’s level of education
would affect their ability to repeat these words.
P5: These words that you’ve said. Only educated peo-
ple would be able to understand it better, to know of
its background. Yes, so it would be easy for him but
for others it’d be a bit of a struggle.
There was consideration for the translated word
‘Constantinople’.
P5: I’ve been a student of history. Constantinople
became Istanbul. So for me, ugh we’ve been study-
ing this since 1960. […] No ordinary person would
know of Constantinople, that what is it? Someone
who’s studied history, yes. That too, religious history.
Islamic history I mean. He would know what is called
Constantinople.
Item 11: language
The phrases were regarded as easy to repeat as they were
commonly known. It was believed average Urdu speakers
should be able to easily complete this item.
P3: They are everyday phrases. Every person uses
them.
P5: This is a common phrase, a common phrase.
Item 12: language
Participant’s perceived images positively and found them
acceptable, easy to recognise, clear of what they repre-
sented and culturally appropriate. They were considered
to be everyday objects that British Urdu speakers would
be familiar with.
P11: Yes, definitely, because all these things are for
everyone. Easy in asking- in looking, in recognizing
because things used in the kitchen, things put on
shelves, for reading, in houses, the random animals,
these with ease.
P19: You see these things everywhere. Like the pea-
cock. Everyone here knows this and they know this.
And I know this. I recognised them, I understood.
Just this one- I couldn’t remember the word tortoise.
Tortoise. No- no- I thought this- this I couldn’t get the
word out for it.
Yet, despite the approval of these images, certain aspects
were mentioned for further improvement by the first 21
participants regarding the book and the suitcase.
P11: To me I found them clear, um for the book a
bit- like a folder and a book, that looks a bit confusing
[…] make it a bit more obvious. And the suitcase of
this is a bit confusing, like a briefcase.
P22: This bag, the way this bag is drawn. It’s not prop-
er. It looks like a brief case type or is it a ladies small
bag.[…] This- it doesn’t look right.
Eyesight was considered as a variable that may affect
performance on this item, so it was suggested the images
should be bigger.
These changes encompassed the updated version
administered to the last four participants, who deter-
mined the images were easier to see with no ambiguity.
Item 13: language
There was a consensus that British Urdu speakers should
be able to answer the questions without issue.
P11: Everyone knows about a camel. This is a desert
animal and everyone knows of this, that it walks with
a big shell, and ugh the one related to the head that
becomes the hat. Everywhere, wherever you go what-
ever the culture may be this is multicultural- hat- mul-
ticultural hat.
However, some indentified how aspects of these ques-
tions may not be known to general British Urdu speakers
and how some questions may cause ambiguity, with the
possible solution of multiple answers.
P5: That’s common. 100% would know. Because ev-
eryone takes a suitcase on travels. […] Has a shell…
yes, this is a little difficult because only tortoises have
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

9
MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
shells. […] Everyone knows about a tortoise but not
everyone knows that it has a shell.
P14: Camel is all over the history. Used camels for
travelling. […] Nowadays, it’s suitcase. But in my gen-
eration before it was camel. […] So people who are
75 or 80 they might start with camel, like car.
Therefore, appropriate changes were made after the
first 21 participants and the updated versions applied to
the last four participants, who were able to answer the
questions without issue.
Item 14: language
The words were considered to be well known words that
Urdu speakers would be familiar with.
P2: Yes. Definitely, will read in the correct way.
P24: Those are straightforward easy words. There’s
nothing difficult about them.
Item 15a: visuospatial abilities; item 15b: visuospatial abilities
Participants understood tasks, and it was the attempt at
drawing the image they found difficult.
P3: In drawing- ugh a person may be weak but they
recognise […] Doing drawing it’s been a long time
having left drawing. […] What I mean is I haven’t
drawn very much. It’s been so long.
P11: For me, to double this number 8 I did not face
any difficulty but for me- to make this box into a 3D
picture, for me was a little- I felt that, how do I make
the lines?
Item 17: visuospatial abilities
Majority of participants found the blacked out letters
easily recognisable to themselves and for average Urdu
speakers. In fact, participants claimed this task was too
easy and measures should be taken to complicate it
further.
P3: They should be a bit more blurry […] take out
another dot, take out a dot here, make it a bit more
difficult. It’s very easy. Because the shape of these
were as the letters are and there was a little change
made in them. Recognising them was not difficult at
all.
DISCUSSION
Our cognitive interviews assessed the understanding and
acceptability of the ACE-III Urdu among British Urdu-
speaking older adults, and the reception to this ACE-III
Urdu was positive. Participants found it easy, straightfor-
ward and culturally acceptable with simple and unambig-
uous language. Majority of items were comprehensible,
and participants understood the tasks being asked. None
of the items were deemed offensive or inappropriate.
However, there were reoccurring issues with five items,
such as 6, which was difficult for the first 21 participants
consistently. The original address was too difficult to
pronounce and understand and not commonly heard of,
which affected the participant’s ability to repeat, recall
and recognise it. Participants also had issues with the
images in item 12 and scored low on item 13 due to the
questions being perceived as vague, with the possibility of
multiple answers.
Therefore, after conducting 21 interviews and observing
these continuous perceptions of these five items being
ambiguous or not understandable, we adapted them
according to participant feedback before resuming inter-
views with the remaining participants who found them
easy to understand, retain and repeat, suggesting the
later version was improved.
Visuospatial abilities tasks, particularly items 15a–15c,
were considered difficult by majority of participants
across all 25 cognitive interviews. However, participants
stated they understood the task, but performance was
low due to drawing ability. Negative responses to the task
were not attributed to the quality of our translation or
cultural adaptation. The same response was seen with
item 3, which required participants to conduct the ‘Serial
7’s’ task. It was understood by all and deemed culturally
acceptable, but the nature of the task was regarded as
complex.
A summary of cognitive interviews determined the
ACE-III Urdu was acceptable with regards to cultural
context, but a few changes would have to be made to
ensure understanding. Therefore, we adapted the ACE-III
Urdu in accordance with feedback, resulting in our final-
ised version for now.
We acknowledge limitations, first in the nature of
our sampling and participant selection. A convenience
sample could not determine that our sample was statis-
tically representative of our target population, and we
were unable to incorporate a more thorough screening
of cognitively healthy participants as there is a paucity of
psychometrically validated cognitive assessments avail-
able for British Urdu speakers. Second, though we were
able to identify a pattern of difficulty in five items after
21 interviews and address these, this should ideally have
been recognised earlier and applied halfway through the
sample instead. Third, our ACE-III Urdu, while culturally
validated, cannot immediately be implemented into prac-
tice until it undergoes a psychometric validation. This
would be the next and final stage of the ACE-III Urdu’s
development, within a British Urdu-speaking population
with dementia versus healthy controls.
Despite this, we were still able to demonstrate a robust
cultural validation technique and demonstrated that it
is not time consuming, easy to conduct and required no
specialist expertise. Our participants were also a good
mix of age and socioeconomic and educational back-
grounds, split across gender. We also ensured thematic
saturation.25
Our research was also conducted in a culturally sensitive
manner that incorporated engagement with the relevant
communities, matched researcher attributes to the target
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from

10 MirzaN, etal. BMJ Open 2018;8:e021057. doi:10.1136/bmjopen-2017-021057
Open access
population, incorporated cultural sensitivity training and
provided materials in English and Urdu to accommodate
for language preference. This illustrates the measures
that can be undertaken to ensure cultural sensitivity of
research undertaken within ethnic populations.
Overall, we detailed the qualitative assessment proce-
dure needed to ensure cultural validation of cognitive
assessments that have been adapted for various ethnic
populations, indicating user involvement through the
cognitive interviewing process. This can allow for cultural
validation to be considered as an essential undertaking
before conducting psychometric validation of any trans-
lated and culturally adapted assessment.
Acknowledgements Nadia Siddiqui and the Pakistani Community Centre,
Manchester.
Contributors All authors made substantial contributions to the conception or
design of the work, or the acquisition, analysis or interpretation of data. NM and
WW drafted the work and revised it critically for important intellectual content.
NM and WW designed all materials and conducted recruitment of participants.
NM underwent training for cultural sensitivity and recruiting ethnic minorities. NM
administered the ACE-III Urdu, conducted all cognitive interviews and their analysis
and write up. WW and NM contributed to the write up of the manuscript. WW and
MP reviewed the manuscript. All authors made the nal approval of the version
published and agreed to be accountable for all aspects of the work.
Funding The authors have not declared a specic grant for this research from any
funding agency in the public, commercial or not-for-prot sectors.
Competing interests None declared.
Patient consent None declared.
Ethics approval This research was approved by the University of Manchester
Research Ethics Committee (UREC Number: 16265).
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data available.
Open access This is an open access article distributed in accordance with the
Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which
permits others to distribute, remix, adapt, build upon this work non-commercially,
and license their derivative works on different terms, provided the original work is
properly cited, appropriate credit is given, any changes made indicated, and the use
is non-commercial. See: http:// creativecommons. org/ licenses/ by- nc/ 4. 0/.
REFERENCES
1. Prince M, Wimo A, Guerchet M, et al. The global impact of dementia:
an analysis of prevalence, incidence, cost and trends. Alzheimer’s
Disease International 2015:1–81.
2. Prince M, Bryce R, Albanese E, et al. The global prevalence of
dementia: a systematic review and metaanalysis. Alzheimers Dement
2013;9:63–75.
3. Prince M, Knapp M, Guerchet M, et al. Dementia UK update.
Alzheimer Society 2014:1–79.
4. Department of Health. Living well with dementia: a national dementia
strategy: Department of Health, 2009:1–100.
5. Chui HC, Gatz M. Cultural diversity in Alzheimer disease: the
interface between biology, belief, and behavior. Alzheimer Dis Assoc
Disord 2005;19:250–5.
6. Richards M, Brayne C, Dening T, et al. Cognitive function in UK
community-dwelling African Caribbean and white elders: a pilot
study. Int J Geriatr Psychiatry 2000;15:621–30.
7. Ofce for National Statistics, National Records of Scotland, Northern
Ireland Statistics and Research Agency. 2011 Census aggregate
data. edn: UK Data Service, 2016.
8. La Fontaine J, Ahuja J, Bradbury NM, et al. Understanding dementia
amongst people in minority ethnic and cultural groups. J Adv Nurs
2007;60:605–14.
9. Black TD. Asian and minority ethnic communities and dementia –
where are we now? Better Health Brieng 2013;30:1–11.
10. Gholap N, Davies M, Patel K, et al. Type 2 diabetes and
cardiovascular disease in South Asians. Prim Care Diabetes
2011;5:45–56.
11. Regan JL. Ethnic minority, young onset, rare dementia type,
depression: a case study of a Muslim male accessing UK dementia
health and social care services. Dementia 2016;15:702–20.
12. Weimer DL, Sager MA. Early identication and treatment of
Alzheimer’s disease: Social and scal outcomes. Alzheimer’s &
Dementia 2009;5:215–26.
13. Parker C, Philp I. Screening for cognitive impairment among
older people in black and minority ethnic groups. Age Ageing
2004;33:447–52.
14. Khan F, Tadros G. Complexity in cognitive assessment of elderly
british minority ethnic groups: cultural perspective. Dementia
2014;13:467–82.
15. Tuerk R, Sauer J. Dementia in a Black and minority ethnic population:
characteristics of presentation to an inner London memory service.
BJPsych Bull 2015;39:162–6.
16. Fillenbaum G, Heyman A, Williams K, et al. Sensitivity and specicity
of standardized screens of cognitive impairment and dementia
among elderly black and white community residents. J Clin
Epidemiol 1990;43:651–60.
17. Lindesay J, Jagger C, Hibbett MJ, et al. Knowledge, uptake and
availability of health and social services among Asian Gujarati and
white elderly persons. Ethn Health 1997;2:59–69.
18. Ganguli M, Hendrie HC. Screening for cognitive impairment and
depression in ethnically diverse older populations. Alzheimer Dis
Assoc Disord 2005;19:275–8.
19. Hsieh S, Schubert S, Hoon C, et al. Validation of the addenbrooke’s
cognitive examination iii in frontotemporal dementia and Alzheimer’s
disease. Dement Geriatr Cogn Disord 2013;36:242–50.
20. Cheung G, Clugston A, Croucher M, et al. Performance of three
cognitive screening tools in a sample of older New Zealanders. Int
Psychogeriatr 2015;27:981–9.
21. Mallinson S. Cognitive interviewing survey validation: A Note.
Lancaster University 2002:1–3.
22. Beatty PC, Willis GB. Research synthesis: the practice of cognitive
interviewing. Public Opinion Quarterly 2007;71:287–311.
23. DeMaio TJ, Rothgeb J, Schwarz N, et al. Cognitive interviewing
techniques: In the lab and in the eld. Answering questions:
methodology for determinig cognitive and communicative processes
in survey research. San Francisco: Jossey-Bass, 1996.
24. Mirza N, Panagioti M, Waheed MW, et al. Reporting of the translation
and cultural adaptation procedures of the Addenbrooke’s Cognitive
Examination version III (ACE-III) and its predecessors: a systematic
review. BMC Med Res Methodol 2017;17:141.
25. Liamputtong P, Ezzy D. Qualitative research methods. Melbourne:
Oxford University Press, 1999.
26. Waheed W, Hassan R. Developing and evaluating a training
programme to increase recruitment of ethnic minority participants
to mental health research: SHAMIL. Poster session presented at: a
lifespan approach. Birmingham, UK: Royal College of Psychiatrists
Faculties of Child and Adolescent Psychiatry and General Adult
Psychiatry, Joint Annual Scientic Conference, 2016:6–7.
27. Gerber ER, Wellens TR. Perspectives on pretesting :“Cognition” in
the cognitive interview? Bulletin of Sociological Methodology/Bulletin
de Méthodologie Sociologique 1997;55:18–39.
28. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative
Research in Psychology 2006;3:77–101.
on 14 December 2018 by guest. Protected by copyright.http://bmjopen.bmj.com/BMJ Open: first published as 10.1136/bmjopen-2017-021057 on 14 December 2018. Downloaded from