Introduction
Very little known is about the exact prevalence of dementia in Pakistan (Hussain et al., Reference Hussain, Rasul, Anwar, Sohail, Kamran, Baig, Shabbir and Iqbal2017). Low- and middle-income countries (LMICs), which includes Pakistan, are projected to show the greatest increase in dementia prevalence over the coming decades (Prince et al., Reference Prince, Wimo, Guerchet, Ali, Wu and Prina2015), largely driven by the rapid increase of life expectancies. However, this is only part of a more complex picture of dementia in LMICs. For example, people within LMICs such as Pakistan are more susceptible to cardiovascular disease and diabetes (Misra et al., Reference Misra, Tandon, Ebrahim, Sattar, Alam, Shrivastava, Narayan and Jafar2017), which subsequently is associated with increased risk of dementia (Whitmer et al., Reference Whitmer, Sidney, Selby, Johnston and Yaffe2005; Kloppenborg et al., Reference Kloppenborg, van den Berg, Kappelle and Biessels2008). Furthermore, under-resourced education systems mean that many LMIC populations tend to be poorly educated, which also is a risk factor of dementia (Livingston et al., Reference Livingston, Sommerlad, Orgeta, Costafreda, Huntley, Ames, Ballard, Banerjee, Burns, Cohen-Mansfield, Cooper, Fox, Gitlin, Howard, Kales, Larson, Ritchie, Rockwood, Sampson, Samus, Schneider, Selbæk, Teri and Mukadam2017).
Pakistan is currently the sixth most populous country in the world, with an estimated older population of 12.5 million in 2017. In percentage terms, older people constitute only about 6.5–7 per cent of the total population, but they are by no means an insignificant segment of the population in terms of their number now. Pakistan belongs to a group of only 15 countries worldwide that have more than 10 million older people. In the future, the rising life expectancy will lead to an increasing number of old and very old people in Pakistan. By 2050, the number of older persons living in Pakistan will reach a staggering 40 million (Zaidi et al., Reference Zaidi, Stefanoni and Khalil2018: 2). Pakistan ranks extremely low in the Global AgeWatch Index, at 92 out of 94 countries, which measures wellbeing of older populations using multi-dimensional indicators. The country ranks particularly low with respect to the health of older persons, with a relatively low life expectancy and even lower healthy life expectancy within the region (Zaidi, Reference Zaidi2013). Despite the rising number of older people in Pakistan and their low socio-economic status, there is very little research on issues linked with old age and dementia (Khan, Reference Khan2014). Greater bodies of dementia literature exist in countries geographically close to Pakistan (i.e. India), however, it is difficult to make inferences due to Pakistan's unique historical, cultural and political background. There is currently no government policy in Pakistan to provide specialist health care for people living with dementia, relying instead on family care-givers playing a central role (Zaidi et al., Reference Zaidi, Willis, Farina, Balouch, Jafri, Ahmed, Khan and Jaffri2019).
Within LMICs such as India, features of dementia appear to be widely recognised (Cohen, Reference Cohen1995; Patel and Prince, Reference Patel and Prince2001). Despite this, dementia is often incorrectly considered to be part of normal ageing, or there are stigmatising beliefs about dementia more broadly (Patel and Prince, Reference Patel and Prince2001; Li et al., Reference Li, Fang, Su, Liu, Xiao and Xiao2011; Maestre, Reference Maestre2012; Ferri and Jacob, Reference Ferri and Jacob2017). Lack of awareness and stigma of dementia can lead to people not to seeking a diagnosis, support, treatment or information about dementia (Iliffe et al., Reference Iliffe, De Lepeleire, Van Hout, Kenny, Lewis and Vernooij-Dassen2005; Alzheimer's Society, 2008; Milne, Reference Milne2010). Therefore, raising awareness of dementia is a worldwide priority, as set out in the World Health Organization's global action plan on dementia (World Health Organization, 2017b), in part because raising awareness is a key pathway to enhance timely diagnosis and reduce stigma (Mukadam and Livingston, Reference Mukadam and Livingston2012). However, before this can be initiated, it is important to establish what people's attitudes, beliefs and knowledge currently are in LMICs, whilst being aware that they will be country and culturally specific. To date, there are limited data originating from Pakistan. One study, carried out more than five years ago, has explored such perceptions in Pakistan, albeit from the perspective carers of people with dementia. In the study, carers were found to have limited awareness of dementia despite their caring role, and that they often believed it was caused as a part of te normal ageing process or secondary to traumatic events or stressors (Qadir et al., Reference Qadir, Gulzar, Haqqani and Khalid2013). Similar findings have been found in Pakistani communities living abroad (Nielsen and Waldemar, Reference Nielsen and Waldemar2016).
The aim of this study is, therefore, to explore perceptions of dementia, its treatment and care in a small group of adult Pakistanis residing in two urban centres.
Methods
Design
Focus group discussions (FGDs) were used to facilitate interactions and are recommended for use to explore beliefs about health and disease (Bowling, Reference Bowling2014). Vignettes were used during the FGDs as a means to depersonalise the content and, therefore, better allow participants to express their views on a sensitive topics (Barter and Renold, Reference Barter and Renold1999; Schoenberg and Ravdal, Reference Schoenberg and Ravdal2000).
Participants
A total of 40 participants (20 male and 20 female) were recruited across two sites in Karachi and Lahore, Pakistan. Situated in the provinces of Sindh and Punjab, respectively, sites were selected because they represent the two most populous cities in Pakistan. Potential participants were identified through word of mouth, local community groups and existing contacts. Participants were selected to maximise heterogeneity in the sample. Participants were excluded if they were currently caring for someone with dementia, as the aim was to explore perceptions of people who had no personal experience of dementia. There was no upper restriction on the age of the participant. All participants were required to be Pakistani and speak Urdu.
Procedure
Each FGD was composed of ten same-sex participants. All participants were provided with an information sheet and informed consent was obtained. The facilitator began each session by introducing the FGD and describing the purpose of the research, after which a series of ground rules were set. The participants were then asked to introduce themselves and were then verbally presented the first of two vignettes. In addition, the vignettes, written in Urdu, were made available on paper to those who were literate. Participants were then asked a series of semi-structured questions by a facilitator regarding each vignette. For example:
• What do you think is happening to [the person] in the story?
• How do you think [the person]’s family feels about her problems?
• What do you think [the person]’s family should do about it?
Following the discussion of each vignette, the facilitator asked a series of semi-structured questions about dementia more broadly. For example:
• What is dementia?
• What is the cause of dementia?
• Who do you think should be responsible for looking after people with dementia?
After the completion of the FGD, participants were asked to complete a series of questions about their demographics, alongside their previous contact with dementia. Researchers read these questions and supported their completion.
All study materials and discussions were in Urdu, which were subsequently translated to English. The FGDs were completed in August 2017.
Two separate research teams ran the FGDs, one in Lahore and one in Karachi:
(1) The Karachi FGDs were led by a female facilitator. The Karachi facilitator was Pakistani born, and was fluent in Urdu, Sindi, English and Balochi. Two additional researchers, one male, one female, who were also Pakistani born, took notes, and provided additional support when needed. The Karachi FGDs were held in a community setting in Korangi, Karachi.
(2) The Lahore FGDs were led by a female facilitator. The Lahore facilitator was Pakistani born, and was fluent in Urdu, Punjabi and English. Two additional researchers, both males, were present as observers during the discussion. The Lahore FGDs were held in the offices of Alzheimer's Pakistan.
Measures
Demographic information was collected from participants, including age, gender, level of education and measures of socio-economic status. In addition, participants were asked to complete a series of questions about their experiences of dementia, which was a translated and adapted version of an existing dementia level of contact questionnaire (Farina et al., Reference Farina, Hughes, Griffiths and Parveen2019).
Vignettes
Two vignettes were created, one describing someone with mild dementia (see Figure 1) and the other with severe dementia. The vignettes were adapted from those previously used in a study exploring awareness of dementia in minority ethnic groups in the United Kingdom (UK) (La Fontaine et al., Reference La Fontaine, Ahuja, Bradbury, Phillips and Oyebode2007). La Fontaine et al. originally created the vignettes to reflect both the language and experiences from previous FGDs whilst utilising symptoms that fit with diagnostic criteria of dementia. In the present study, the two vignettes were further adapted to ensure that they were appropriate for a Pakistani population. In line with common practice, the vignettes were rewritten in the third person and ages of the cases were added (Randhawa et al., Reference Randhawa, Jiwa and Oremus2015). Specifically, the vignettes used refer to two hypothetical cases of Mrs Shabnam and Mrs Gulnara, both 72 years old. The vignettes were reviewed by British dementia-related researchers, British dementia-related health-care staff, and Pakistani academics and clinicians to ensure that the descriptions were believable and accurately reflect symptoms of dementia and its severity. The vignettes were then translated into Urdu and reviewed by a Patient and Public Involvement Group in Pakistan, who provided further input into the vignettes. The final versions were then independently reviewed by a bilingual researcher to ensure that the translated versions maintained their meaning.
Figure 1. An example vignette: Mrs Shabnam, 72 years old with mild dementia.
Source: Adapted from La Fontaine et al. (Reference La Fontaine, Ahuja, Bradbury, Phillips and Oyebode2007).
Analysis
Demographic information and previous experience of dementia were reported descriptively (e.g. frequencies).
All audio recordings were transcribed in Urdu initially and then translated into English by multilingual researchers in Pakistan. A thematic analysis using inductive approaches was carried out, taking an interpretivist epistemological position. The lead British author of this paper used an open-coding approach to code a single FGD transcript, utilising phrasing as close as possible to the participant's own words. This was then independently reviewed by three researchers (one British, one Irish and one Pakistani) who added and commented on the emerging coding framework; disagreements were resolved through discussion amongst the researchers. The lead author then continued to code the remaining transcripts, adding new codes when identified. Codes reflecting connected concepts were then grouped into higher-order sub-themes and themes.
Ethics and rigour
Informed consent was obtained from all participants. For participants who were illiterate, the information sheet, consent forms and vignettes were read out to them.
To ensure that the Urdu transcriptions were translated accurately, without losing the original meaning, translations were completed by the multilingual research team who led the FGDs. We attempted to keep the wording as close to the direct translation where possible, but sometimes the wording was altered to ensure that it made grammatical sense. In some situations the choice of wording is pertinent to understanding the context; this has remained unchanged but is expanded within brackets.
A single FGD from each site was independently observed by a researcher (NF or AZ) to make notes on FGD procedure and to provide additional rigour. During the coding, the researcher (NF) met with an independent researcher to discuss coding themes and to support the process of reflexivity. NF was able to reflect upon his own position as a researcher within the UK and how this might affect the interpretation of findings. The final coding and formation of themes were discussed amongst members of the research team and broader scientific advisory board (composed of Pakistani clinicians and academics) to ensure the findings remained valid.
Results
Across all sites, 40 participants completed the FGDs. On average, participants were 37.2 years old, lived in a household of 6.5 people and most commonly had been educated to A-level or equivalent Faculty of Science standard (N = 9). For further details of the demographics of each site, see Table 1. The majority of participants had never encountered someone with dementia (N = 30, 81.1%), spoken to family or friends about dementia (N = 27, 73.0%), been taught about dementia in school (N = 35, 94.6%) or watched television shows or movies in which a character has dementia (N = 24, 64.9%).
Table 1. Demographic details of participants across the four focus group discussions
Notes: SD: standard deviation. Rs: Pakistani rupees. FSc: Faculty of Science. HSSC: Higher Secondary School Certificate. SSC: Secondary School Certificate.
From the FGDs, five key themes were identified: (a) dementia awareness, (b) responsibility, (c) religion, (d) barriers to health care, and (e) identified support needs.
Dementia awareness
Participants tended to have very little knowledge of dementia, identifying a variety of causes for the symptoms described in the vignettes. Most prominently, participants felt that the symptoms described in both vignettes were due to normal age-related decline:
Commonly, this is because of old age that it is an ailment. (Female, Karachi, FK1)
Second to normal age-related decline, many participants also incorrectly identified that the cause of dementia was due to life-stresses, often referred to as ‘tension’:
…because of the ‘tension’, the person has the disease of forgetting [dementia]. (Female, Karachi, FK7)
It [dementia] happens because of worries. (Male, Karachi, MK6)
There were a few participants who did understand that the vignettes were about dementia, or more commonly recognised it as being related to ‘forgetfulness’. It is also important to highlight that participants were correctly able to recognise that a diagnosis of diabetes (as described in the first vignette) might be associated with dementia:
…one thing that I noticed is that the forgetting issue is somehow being related to diabetes. (Male, Lahore, ML1)
Irrespective of the cause, misconceptions led to some participants holding incorrect views of prognosis and the treatment for those with dementia. A sub-group of participants believed that symptoms might improve, or the disease is curable:
…if her household and herself takes a little more care … Then her ailment will disappear. (Male, Karachi, MK6)
Responsibility
Responsibility featured heavily throughout the FGDs, with participants commonly identifying that the responsibility of accessing or providing dementia care lies with either the person with dementia or their family. The perception that it was the family's responsibility to provide care was driven, in part, by societal norms. This sometimes led to an emphasis on children taking a greater responsibility compared to other family members:
It is the responsibility of the children to look after their mother and father. God has given us the courage; it is our right [privilege] that we should take care [of them]. (Female, Karachi, FK6)
Tied with this, participants saw the cases described in the vignettes as being the fault of the families because they did not give enough time or attention.
Participants in the FGDs also identified that there was an element of personal responsibility when it came to dementia and its symptoms. More interestingly, some felt that the individual acted as a barrier (e.g. lazy or apathetic) to their own health care:
In my view Shabnam herself was hurdle for her treatment, because she had stopped meeting relatives. She has become fed-up. [So] she was herself a hurdle and nothing else. (Male, Karachi, MK6)
As such, it was seen that the person with dementia is partially or wholly responsible for their own care or treatment:
In my opinion she [Shabnam] should remove these barriers herself. (Male, Karachi, MK7)
Mrs Shabnam should herself take steps; she can bring herself back with her own willpower. She should take steps herself and household members should also push her. (Female, Karachi, FK8)
Religion
Closely related to responsibility, religion was also used to provide context as to why the family members should provide care for someone with dementia, often in very certain terms. Participants usually rationalised that it was the children's duty to care for their parents, often citing religious texts:
It is very clearly mentioned in the Holy Quran that you need to treat your parent like your parents treated you when they were young. And you are not allowed to be mad at them, you can't even utter a single word in anger. From this saying we get the lesson that we should take care of them and never leave them at any stage. We should always give them attention and we shouldn't neglect them. We shouldn't ignore them. This is the basic religious lesson. (Female, Lahore, FL1)
If you follow religion, then you cannot leave your parents until the end. If you are religious, then you should remain loyal [to your parents] until the end. (Male, Karachi, MK1)
Religion was also used as a means of describing how best to treat dementia as well as to inform its prognosis. However, participants tended not to provide reference to religious texts or teachings when describing these opinions. A common theme, particularly amongst the Karachi FGDs, was that prayer could be a means of preventing or treating dementia:
Whoever keeps Allah pleased then, everybody is pleased. If we perform Farz Footnote 1 then, it is my conviction that we all can remain safe from disease. Everybody should pray [and follow their religious values]. (Female, Karachi, FK3)
If we remember Allah then we can remain safe from all the diseases. (Female, Karachi, FK5)
Barriers to health care
Participants often struggled to identify potential barriers to accessing care for someone living with dementia, aside from self- or family-related barriers (see above). However, finances were occasionally identified as a key barrier to accessing support and health care:
Household members would like to [medically] treat her, but they don't have the money. (Male, Karachi, MK1)
Some of the participants also felt that the current health-care system was a barrier to receiving care. Either because of feelings that the doctors were not sufficiently trained, that they did not have the patients’ best interests at heart or that they would make an incorrect diagnosis:
If I take them to a government hospital, then the doctor will tell me that their symptoms are due to old age, therefore I should take care of them and pray for them. (Female, Lahore, FL2)
A smaller group of participants identified that stigma may act as a barrier; either from the stigma derived from receiving a formal diagnosis or the fact that getting external help was seen as a failure on the part of the family:
That if we will send our parents to day care then what will other people say about us? If we will hire some helper then what will the people think or say about us? (Female, Lahore, FL2)
It was only within perceived barriers where the gender of participants appeared to influence responses. Female participants were more likely to highlight that the family dynamics could act as a barrier, with views of certain family members being prioritised over others. For example, if was felt that if the daughter identified a need for additional support this could be overridden by the son.
Identified support needs
Despite individuals’ limited knowledge of dementia, participants identified that more could be done for people with dementia in Pakistan. Most frequently, participants identified the need to raise more awareness about dementia. A portion of participants suggested that television and radio were the best means of raising awareness, though others felt that using a strategy similar to those used to raise awareness of polio and dengue fever (discussed in next section) would be more effective. Whilst participants did not often raise the view that the government should be responsible for dementia care, it was seen as being primarily responsible for raising awareness.
Participants believed that new facilities (e.g. hospitals) and specialists were needed, though this was predicated with the notion that any facilities opened should be accessible, low-cost and relaxing (i.e. not stressful). Most importantly, it was seen that any new facilities and services should be accompanied by information about how and where to access this health care:
Institutions should be opened immediately and the ones that are open should provide environment akin to home. (Male, Karachi, MK2)
Discussion
The purpose of this study was to explore the perceptions of dementia amongst a series of focus groups of Pakistani adults. Through the use of vignettes and FGDs, we were able to identify five key themes that summarised their perceptions of dementia.
The FGDs highlight that there is often lack of knowledge and awareness of dementia amongst Pakistani adults, leading to incorrectly identifying the cause of symptoms, dementia's prognosis and how best to treat it. In line with previous literature from other LMICs, the common view was that dementia is a normal part of ageing (Patel and Prince, Reference Patel and Prince2001; Li et al., Reference Li, Fang, Su, Liu, Xiao and Xiao2011; Ferri and Jacob, Reference Ferri and Jacob2017). This is perhaps unsurprising considering the majority of participants had no experience of dementia during their education or throughout their lives. Lack of awareness of dementia is likely to act as a barrier to accessing support services in the future, with greater knowledge of dementia being associated with increased intention to seek help (Werner, Reference Werner2003). Perhaps more concerning is the fact that some participants associated dementia symptoms with being the fault of the individual (e.g. they are their own hurdle) or family, which is potentially stigmatising. This could explain, in part, why there appears to be little dialogue and research about dementia.
It should be noted that there did appear to be knowledge or recognition of dementia from a sub-group of participants within the study, with participants correctly referring to dementia as a disease of ‘forgetting’. This may not be surprising because previous qualitative evidence from India has shown that the features of dementia are generally widely recognised (Cohen, Reference Cohen1995; Patel and Prince, Reference Patel and Prince2001). Similarly, the term ‘disease of forgetting’ has specifically been used within other LMICs such as Tanzania to describe dementia (Mushi et al., Reference Mushi, Rongai, Paddick, Dotchin, Mtuya and Walker2014). Obviously, forgetfulness does not capture fully what dementia is, but recognition of symptoms is important. Participants’ ability to draw links between dementia symptoms and diabetes may be indicative of some recognition or implicit knowledge of dementia risk factors (Ott et al., Reference Ott, Stolk, Van Harskamp, Pols, Hofman and Breteler1999; Kloppenborg et al., Reference Kloppenborg, van den Berg, Kappelle and Biessels2008).
Raising dementia awareness was seen as a priority by participants. Television and radio were seen as the best way to reach the majority of people, whilst many felt that the use of awareness campaigns similar to those for polio or dengue fever would be beneficial. In the case of polio, a mass media campaign and targeted community engagement was used to raise awareness (National Emergency Operation Centre, 2017), but this was complemented with support from international leading Islamic scholars (World Health Organization, 2014). It has previously been commented that Islamic institutions play an important role in promoting mental health through routes such as public information and health education (Baasher, Reference Baasher2001).
Even in this sample of Pakistanis, with little knowledge or awareness of dementia, participants were able to identify barriers to accessing health care, such as financial restrictions. Whilst wealthier Pakistanis may have private health-care insurance, out-of-pocket payments are the largest source of private expenditure for health care (65% in 2014) (World Health Organization, 2017a). Even government facilities that are officially free of charge may result in individuals giving unofficial payments or bribes due to them being over-utilised and under-resourced (Malik, Reference Malik2015).
It is unsurprising that religious beliefs were used as a means to provide a direction in relation to dementia care, as 96.3 per cent of Pakistanis are Muslim (Pakistan Bureau of Statistics, 2019). Notably, the family (in particular adult children) were seen as being responsible for providing such care, which was tied in with the teachings of the Quran:
And your Lord has decreed that you worship none but Him. And that you be dutiful to your parents. If one of them or both of them attain old age in your life, say not to them a word of disrespect, nor shout at them but address them in terms of honour. (Quran 7:23, Muhsin Khan translation)
As supported by evidence from South Asian family carers in the UK (Mackenzie, Reference Mackenzie2006), religious obligation, and the associated stigma associated with seeking help, is therefore likely to lead to family carers becoming isolated. This is not to say that the views from the FGDs fully capture the complexities in theological interpretations of dementia care within Islam, and it may be that the group setting led participants to avoid conflict and share more socially accepted, normative opinions (Smithson, Reference Smithson, Alasuutari, Bickman and Brannen2008). Irrespective of this, it is evident that an individual's understanding of Islam plays an important role in shaping opinions about dementia. This can be observed in people with a diagnosis of dementia in Pakistan, with some being unclear about the religious obligations of those with dementia (Willis et al., Reference Willis, Zaidi, Balouch and Farina2018). Future research using individual interviews may assist in delving deeper into this complex topic.
There are a number of key strengths of this study. First, the FGDs were composed of participants with a range of education levels, ages and household incomes. The heterogeneity within the FGDs maximises the variability in opinions from these metropolitan areas. Recruiting from two urban centres also reduces the likelihood that the sample reflects the opinions and views of a single region. Second, the FGDs were organised to be homogenous within each group in terms of gender (i.e. single-sex). Creating groups with more homogenous characteristics not only facilitates productive group discussions, but also makes it easier for comparisons to be made between groups based on a single characteristic (Hennink, Reference Hennink2007; Krueger and Casey, Reference Krueger and Casey2014). In particular, gender has previously been identified as being an important characteristic to consider, with the gender composition of a focus group affecting the nature of the interaction (Krueger and Casey, Reference Krueger and Casey2014; Stewart and Shamdasani, Reference Stewart and Shamdasani2014), with females in mixed-sex focus groups tending to be less dominant. Finally, the use of vignettes allowed participants to discuss a potentially difficult topic by depersonalising it (Barter and Renold, Reference Barter and Renold1999; Schoenberg and Ravdal, Reference Schoenberg and Ravdal2000). It also meant that all participants received the same description of symptoms associated with dementia, which provided a standardised context in which to discuss the topic.
The study did, however, have several limitations. The findings reflect the views of a small group of adults from Pakistan. Every effort was made to encourage participation from people with a range of social and economic backgrounds. However, as recruitment was through word of mouth, and the results are derived from a small group, the views are unlikely to reflect fully the views of all 207 million Pakistanis. Importantly, participants were primarily recruited from two metropolitan centres (Lahore and Karachi) and, therefore, may not capture the views from Pakistanis residing in more deprived, rural areas. It is also possible that certain demographics may have chosen not to participate in the FGDs because of their views on the topic, or the setting in which the FGDs were run. For example, in the present sample, 85 per cent of participants achieved over a primary level of education, which is likely to reflect a more educated Pakistani population considering the Gross Enrolment Rate alone to primary schools (2016–2017) was 87 per cent (Government of Pakistan, 2017). As highlighted above, the FGDs also may not fully reflect more extreme views due to group settings encouraging more socially accepted opinions to be shared (Smithson, Reference Smithson, Alasuutari, Bickman and Brannen2008). As such, distributing an anonymous survey of dementia attitudes and knowledge may be a logical next step to provide more generalisable evidence, whilst also being in a format that would be less susceptible to reporting socially acceptable opinions.
This is the first study to explore, in depth, the perceptions of dementia in Pakistani adults who have had no specific contact with people with dementia. Although the findings only reflect a small sample of adults residing in urban centres, it provides a unique insight into common misconceptions about dementia and also areas in which to help individuals to access support. This study is an important first step in the advocacy for dementia care reform in Pakistan. It is evident that raising awareness about dementia is essential; however, this needs to be accompanied by improved specialised dementia health-care services. Religion obviously plays an important role in shaping attitudes within Pakistan, and therefore future efforts should explore how best to shape the narrative surrounding dementia in Islam.
Acknowledgements
A special thank you to all the participants that took part in the research, and the user involvement in Pakistan (Alzheimer's Pakistan). We also wish to thank our project partners in Pakistan, particularly Hussain Jafri of Alzheimer's Pakistan for the data collection in Lahore, and Qurat Khan of the Aga Khan University, and Rubina Jaffri and Irfan Ahmed of HANDS for the data collection in Karachi, without whom this project could not have taken place. We are grateful for the encouragement and support from Chris Roles (Chief Executive Officer of Age International) and Yasmin Rashid (Minister of Punjab for Primary and Secondary Healthcare, and Specialised Healthcare and Medical Education). We are grateful for the contribution of Laura J. Hughes who acted as an independent consultant during the research.
Author contributions
NF wrote the first draft of the paper, helped plan the study, performed the analyses and revised the paper. AZ, RW and SB helped plan the study, revised the paper and contributed to the analysis.
Financial support
This work was supported by Age International, Age UK and HelpAge International.
Conflict of interest
The authors declare no conflicts of interest.
Ethical standards
The University of Southampton and Aga Khan University ethics committees approved this study.
