Objectives Pulmonary hypertension (PH) is a life-shortening disease associated with early mortality and high morbidity. With advancements in medical treatment, people are living longer with the disease, and research is now needed to explore variables that help to enhance patient-reported outcomes. This study investigated the coping strategies of individuals with PH and examined the relationship between coping, depression, health anxiety and health-related quality of life (HRQoL).
Design A cross-sectional survey design was used.
Participants Participants (n=121) were recruited from membership of Pulmonary Hypertension Association (PHA) UK.
Outcome measures Participants completed a series of questionnaires assessing depression (Patient Health Questionnaire-9), health anxiety (Short Health Anxiety Inventory), HRQoL (emPHasis-10) and coping (Brief COPE). A principal component analysis was used to identify participants’ coping profile. A series of correlational, linear and moderated multiple regression analyses were performed to examine the relationship between coping and health-related outcomes.
Results Overall, 43% participants met criteria for potential clinical depression. Depression and health anxiety were strongly associated with HRQoL, explaining 37% and 30% of variance respectively (p<0.001). A principal component analysis identified a four-component model of coping. Dimensions were named based on construct items: ‘cognitive and affirmation coping’ (seven items), ‘passive coping’ (four items), ‘external coping’ (seven items) and ‘substance use coping’ (two items). Cognitive and affirmation and external coping moderated the relationship between depression and HRQoL, with high use of these coping strategies reducing the impact of depression on HRQoL. External coping also moderated the effect of health anxiety on HRQoL.
Conclusions The results uniquely highlight the importance of coping styles and psychological distress in predicting HRQoL in PH. Our findings indicate the importance to assess for psychological distress in this population and suggest the need to offer psychological interventions that take into account coping resources and strategies.
- mental health
- anxiety disorders
- respiratory medicine (see thoracic medicine)
Data availability statement
Data are available on reasonable request.
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/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
Understanding how patients cope with the demands of pulmonary hypertension (PH) can be used to help inform services and patient-reported outcomes.
The results demonstrate high rates of distress in PH within a community sample, with 43% of people scoring above the clinical cut-off for depression.
Coping strategies were found to moderate the relationship between health anxiety and depression, and health-related quality of life (HRQoL) in PH.
Coping by seeking external support and cognitive-based strategies buffered the impact of psychological distress on HRQoL.
Clinical factors, such as PH severity, were not included as covariates, and therefore, it is not clear how objective measures of the impact of PH influence coping.
Pulmonary hypertension (PH) is a progressive disease characterised by elevated pulmonary artery pressure, which ultimately leads to right heart failure and death if untreated.1 PH is a global term used to describe five distinct groups of the disease: group 1: pulmonary arterial hypertension (PAH); group 2: PH due to left heart disease; group 3: PH caused by lung disease; group 4: chronic thromboembolic PH; and group 5: PH due to miscellaneous causes.2 Common symptoms of PH can include shortness of breath, fatigue, pain, weakness, dizziness, oedema and in some cases syncope.3
Better disease awareness and a greater understanding of the aetiologies of PH have helped to advance diagnostic and therapeutic approaches, dramatically improving treatment outcomes, and resulting in lower mortality rates.4 5 As in other conditions in which people are living longer with a disease, research is now needed to examine the impact on daily life and identify variables associated with patient outcomes such as quality of life, the evidence from which can be used to inform care and support.6
The WHO functional class can be used to help indicate the severity of patients’ symptoms of PH. This index ranges from I, indicative of patients experiencing none to mild PH-related symptoms during physical activity, to IV suggesting patients will experience severe symptoms at rest.7 However, the WHO classification system was not designed to reflect the psychosocial impact of living with PH. A qualitative synthesis of adults’ lived accounts summarised the specific challenges posed by the disease, including the burden of living with disabling symptoms, intensive treatment regimens, impact on loved ones, stigma of PH and lack of understanding from others.8
Coping has been viewed as cognitive, emotional and behavioural strategies employed in an attempt to manage distress perceived as threatening.9 This response requires the individual to recognise and appraise the stressor, taking into consideration the expected outcome and their perceived ability to cope.10 A large evidence base exists examining how people cope with living with long-term health conditions11–14; however, no studies to date have quantitively investigated coping in people with PH. Understanding how patients cope with a disease can help to identify targets for intervention and assist in identification of individuals most at risk of experiencing difficulties.15 This is particularly relevant in PH, as treatments aim to sustain health-related quality of life (HRQoL), and early intervention is viewed as essential in reducing morbidity.5
The primary aim of this study was to explore the specific coping strategies used by people with PH. We used the Brief COPE measure,16 which explores 14 different coping strategies, including: self-distraction, denial, substance use, behavioural disengagement, venting, self-blame, active coping, use of emotional support, use of instrumental support, positive reframing, planning, acceptance, humour and religion. This tool is one of the most commonly used measures to assess coping in clinical populations.17 While some studies have grouped coping strategies dichotomously (ie, approach-focused and avoidance-focused coping strategies), there is no clear consensus on scales underlying composition across samples. Moreover, given that the questionnaire was not designed for this purpose, the developer suggests that researchers should group items based on datasets collected from specific samples as the pattern of relationships between items will differ between groups of individuals.18 Indeed, the variation in subscales identified across clinical samples is likely to reflect the idiosyncratic challenges associated with specific life events. In fact, numerous studies using this measure have applied different statistical analyses to examine the psychometric properties of higher order structures,19 including principal component analyses to combine items to identify components that define coping styles components.20 21
A secondary aim of the current study was to assess how coping strategies are related to depression, health anxiety and HRQoL, using a series of correlational and regression analyses. As observed elsewhere, the relationship between psychological factors and burden of a long-term health condition is likely to be influenced by the coping strategies an individual uses (Hundt et al 2015). Lower HRQoL,22 and depression and anxiety are commonly reported as high in people with PH.22–24 Although health anxiety has not specifically been examined in this condition, there is evidence to suggest individuals experience anxiety about their condition, which could further contribute to distress.8 Therefore, we hypothesised depression and health anxiety would be related to lower HRQoL, and this relationship would be moderated by specific components of coping strategies.
A cross-sectional design was employed, using a convenience sampling method. The article has been conducted in line with the Strengthening the Reporting of Observational Studies in Epidemiology checklist (online supplemental file 1).25
Nahlen Bose et al26 examined the relationship between HRQoL and coping in individuals with chronic heart failure. Coping significantly predicted the physical (R2=0.26, p<0.001) and mental (R2=0.4, p<0.001) health composite of the QoL measure, equating to an observed effect size (f2) of 0.35–0.67. A prior analysis using G*Power, with an effect size f2 of 0.35, alpha level of 0.05, 80% power suggested a total sample size of at least 25 was required.
Participants were recruited over the span of 2 weeks from advertisements electronically sent to members of Pulmonary Hypertension Association (PHA) UK. To be eligible, participants must have been aged 18 years or over, living in the UK, diagnosed with PH, able to complete self-report questionnaires independently in English and provide informed consent.
Individuals interested in the study were instructed to access a website link directing them to the participant information sheet. If individuals still wished to take part after reading the information sheet, they were asked to complete a consent form and series of questionnaires (see further). All measures were hosted by Qualtrics.
This included name, age, gender, ethnicity, living status, employment status, level of education, PH diagnosis and functional class, and duration of PH.
Quality of life
The emPHasis-10 is a PH specific HRQoL measure.27 Participants are asked to rate how 10 problems related to life with PH have affected them ‘recently’ on a unidimensional scale. Possible scores range from 0 to 50, with a higher score indicative of lower HRQoL. The measure has excellent internal reliability (α=0.9)27; Cronbach’s alpha within the current study was α=0.94.
The Patient Health Questionnaire-9 (PHQ-9)28 was used to measure symptoms of depression. This is a nine-item measure asking participants how often they have been bothered by a range of depression-related difficulties over the previous 2 weeks. Scores range from 0 to 27; a score of 0–4 indicating ‘no difficulties with depression’, 5–9 ‘mild difficulties’, 10–14 ‘moderate’, 15–19 ‘moderately severe’ and 20 or higher ‘severe depression’. A cut-off of ≥10 has been found to be a valid and reliable measure for major (clinical) depression.29 This measure has previously been used in those with PH.30 The PHQ-9 was found to have good internal consistency (α=0.88) in the current sample.
The Short Health Anxiety Inventory (SHAI)31 was used to measure participants’ anxiety in response to how they appraised their health. This measure asks respondents to endorse items assessing their thoughts, feelings and behaviours about health anxiety over the past week on a scale ranging from 0 to 3 according to specific statements. While the SHAI is typically an 18-item measure, it can be converted to 14 items when exploring health anxiety in people with an existing medical disease. The 14-item measure has been found acceptable in clinical samples, demonstrating high internal consistency (Cronbach’s alpha 0.84).32 There is not a cut-off point that has been used consistently in the literature; however, one study performed a discriminant function analysis that suggested a score of 18 or higher identified people with significant health anxiety (hypochondriasis).33 Internal consistency was excellent (α=0.9) in this sample.
The Brief COPE measure16 was used to examine participants’ coping. Individuals were asked how often they engage in 28 behaviours when experiencing hardship in their life, which are aggregated into 14 coping strategies (two items per each coping strategy). Possible scores for each strategy ranged from 2 to 8, with a higher score indicating greater use. Cronbach alphas were all within the acceptable range or higher (α=0.65–0.94), with the exception of denial (α=0.58).
HRQoL scores were normally distributed (Kolmogorov-Smirnov p=0.085); therefore, parametric tests were used. Descriptive statistics were used to describe participants’ sociodemographic and clinical characteristics, and coping profile. No data were missing.
A principal components analysis was used. A direct oblimin rotation was applied as there is a strong theoretical basis for the coping strategies investigated here to be related.34 The determinant of the correlation matrix was 1.379, suggesting this approach was suitable. Solutions were analysed using Kaiser-Meyer-Olkin measure of sampling adequacy, Bartlett’s test of sphericity, communalities, scree plot, variance explained and the component matrix. Missing data were excluded using pairwise deletion.
A series of Pearson’s correlation tests were performed to explore the association between HRQoL and self-reported depression, health anxiety and coping strategies. To test whether depression or health anxiety predicted HRQoL and explore whether different coping styles moderated this relationship, a series of moderated multiple regression analyses were conducted—all predictor variables were mean-centred. Significant interaction terms were decomposed using a series of simple slope analyses; which involved seperating coping strategies from continious variables into dichotomous variables. This enabled us to examine the relationship at high (1SD above the mean) and low (1SD below the mean) use of each coping strategy. Stepwise multiple regression analyses were performed to identify whether coping components significantly predicted scores of psychological distress (depression and health anxiety). It is not common practice to correct alpha levels in regression studies and so the alpha level of 0.05 was accepted.35 All data analyses were performed using SPSS V.25.
Patient, public involvement
The conception and research design of the study was developed in collaboration with PHA UK.
Sociodemographic and clinical characteristics
Demographic characteristics of the sample (n=121) are summarised in table 1. On average, participants had PH for 8.8 years; however, a large variance was observed. Most participants had either idiopathic PH or chronic thromboembolic PH; however, 13.2% were not sure of the group of their diagnosis. The largest group of individuals did not know their functional class (39.7%); of those who did, class III and II were most reported.
Participant’s mean HRQoL was 23.6 (SD=13.5) and health anxiety 15.4 (SD=7.3). On average, participants scored 8.9 (SD=6.2) on the measure of depression. Findings suggested less than one-third experienced no symptoms of depression (29.8%), whereas 42.9% of the sample met criteria indicative of the possible presence of clinically significant depression warranting the need for intervention (see table 1).
Participants’ scores on each of the 14 strategies of coping are displayed in table 2. Overall, the most frequently used coping strategy was acceptance and self-distraction, whereas substance abuse and behavioural disengagement were the least common.
Results from the principal component analysis of the 28 Brief COPE items, indicated the Kaiser-Meyer-Olkin measure of sampling adequacy was 0.75 (above the recommended value of >0.6). Bartlett’s test of sphericity was significant (χ2 (378)=1719.2, p<0.001). Communalities for the 28 items were >0.3. Therefore, findings support the use of the proposed method to examine data collected using the Brief COPE.
The initial eigenvalues and scree plot indicated eight components (>1 eigenvalue) accounted for 70.5% of the cumulative variance; the first four components accounted for 25.1%, 10.3%, 9.2% and 7% (>2 eigenvalue), whereas the remaining four explained 5.3%, 5.1%, 4.5% and 3.9% (<1.5 eigenvalue). A four-component solution was preferred because of the levelling off of variance as indicated by eigenvalues and visual inspection of the scree plot (figure 1), which demonstrated a discontinuity in the curve after the fourth components. Moreover, many of the items that loaded (>0.4) on components five to eight were also coloaded (>0.3) on the first four components.
The final four components model explained 51.6% of the cumulative variance (all communalities remained at >0.3). After inspecting the pattern matrix, eight items were removed because they coloaded (>0.3) on more than one component, which were related to active (n=2), denial (n=2), humour (n=2), planning and venting; thus, no items relating to denial, active coping or humour were included in the final model. All 28 items loaded on at least one primary component (>0.4).
The first component included items relating to distraction (n=2), positive reframing (n=2), acceptance (n=2) and planning (n=1). Cronbach’s alpha was α=0.8. As these strategies reflect participants’ thinking processes, the component was named ‘cognitive and affirmation coping’. The second component consisted of items assessing behavioural disengagement (n=2) and self-blame (n=2). Cronbach’s alpha was α=0.7. These strategies are characteristic of avoidance and unhelpful approaches; as such, this component was labelled ‘passive coping’. The third component represented items assessing emotional support (n=2), instrumental support (n=2), religion (n=2) and venting (n=1). Cronbach’s alpha was α=0.8. Given all these strategies involve seeking support from others, it was termed ‘external coping’. The final component included items addressing substance use (n=2). Cronbach’s alpha was α=0.9; this construct was subsequently named ‘substance use coping’. Composite scores were created for each strategy of coping by summing the items; a higher score indicating greater use (table 3).
Relationship between coping style, depression, health anxiety and HRQoL
A series of Pearson’s correlational analyses demonstrated HRQoL was significantly related to health anxiety (r=0.55, p<0.001) and depression (r=0.61, p<0.001). In turn, health anxiety and depression were significantly correlated (r=0.65, p<0.001).
Regarding coping strategies, passive coping was significantly positively correlated with depression, health anxiety and lower HRQoL. Substance use was positively related to depression and passive coping. External coping was positively associated with health anxiety and cognitive and affirmation coping. Cognitive and affirmation coping was not significantly related to any clinical measure (table 4). A stepwise multiple regression analyses revealed passive coping significantly predicted both depression (F(1,118)=103.79, p<0.001) accounting for 46% of the variation, and health anxiety (F(1,118)=43.53, p<0.001) accounting for 26%; the other approaches were excluded by the analysis.
A linear regression analysis revealed depression significantly predicted HRQoL, accounting for 37% of the variance (F(1,119)=70, p<0.001). A series of moderated regression analyses were performed by adding in the interaction term between coping strategy (cognitive and affirmation coping, passive coping, external coping and substance use coping) and depression (table 5). The relationship between depression and HRQoL (ΔR2=0.4, ΔF(3,116)=26.1, standardised β=0.63, p<0.001) was found to be moderated by cognitive and affirmation coping (standardised β=−0.18, p=0.017) as indicated by significant interaction effects. The relationship between the same variables (ΔR2=0.4, ΔF(3,116)=25.2, standardised β=0.62, p<0.001) was also moderated by external coping (standardised β=−0.14, p=0.05). No significant interaction effects were observed in passive coping or substance use. The interactions were decomposed using simple slopes analyses, which demonstrated there was a buffering effect of both cognitive and affirmation coping and external coping strategies, such that high levels of cognitive and affirmation (standardised β=0.47, p<0.001) and external coping (standardised β=0.48, p<0.001) weakened the relationship between depression and HRQoL. Conversely, low levels of coping strengthened the association (standardised β cognitive and affirmation=0.79, p<0.001; standardised β external=0.76, p<0.001).
Health anxiety was also found to significantly predict HRQoL, explaining 30% of the variance (F(1,119)=51.1, p<0.001). The same analysis described previously suggested the relationship between health anxiety and HRQoL (ΔR2=0.35, ΔF(3,116)=20.45, standardised β=0.58, p<0.001) was moderated by external coping (standardised β=−0.17, p=0.03), as indicated by significant interaction effects (table 5). A simple slopes analysis revealed high use of external coping reduced the association between health anxiety and HRQoL (standardised β=0.42, p<0.001), whereas low use increased the relationship (standardised β=0.74, p<0.001).
This is the first study to investigate the coping profile of people with PH. Using the Brief COPE tool, we identified a four-component model of coping in PH, and we have investigated how these four coping styles relate to depression, health anxiety and HRQoL. Our findings are consistent with other research demonstrating the high rates of depression and anxiety in those with PH.23 However, given that the current sample was recruited from the community, as opposed to individuals under the care of services, it is alarming that 42.9% of participants met the cut-off indicative of the possible presence of clinically significant depression and 34.7% for health anxiety. Moreover, to our knowledge, this is the first study examining anxiety relating to health specifically in PH. On the group level, participants’ scores (M=15.4) were greater than what has been reported using the same measure in patients with a diagnosed anxiety disorder (M=14.9); individuals attending gastroenterology (M=11.4), general practitioner appointments (M=11.2) and MRI appointments (M=10.6); students (M=9.6); and non-patient controls (M=9.4).31 The high levels of depression and anxiety found in our community sample are at a stark contrast to the fact that just over one-third of patients with PH who present with mood difficulties receive treatment.36
People with PH demonstrated employing a range of strategies to cope with life adversities, with a greater tendency to use acceptance, self-distraction and active coping, as opposed to substance abuse and behavioural disengagement. Results indicate this is likely to be a helpful approach, as our analysis demonstrated that the latter two coping strategies loaded on two of the four components—passive coping and substance use coping—which were related to depression, health anxiety and lower HRQoL. In fact, in a stepwise regression analysis, passive coping was the only significant predictor of depression and health anxiety suggesting that although these coping strategies are used less frequently in this group, it is important to identify individuals who are at risk of engaging in such behaviours. However, it should also be noted that due to the cross-sectional nature of this study, we cannot make any strong conclusions regarding the direction of the relationship, for example, whether passive coping is a causal factor or a symptom of depression or anxiety.
The third coping strategy identified in the analysis was termed external coping, characterised by behaviours through which participants sought practical, emotional and spiritual support from others. This coping strategy was found to be an important moderator, mitigating (although not fully eliminating) the impact of both depression and health anxiety on HRQoL. Receiving support from others is a common strategy of coping observed across populations living with a long-term condition.37 However, patients with PH have previously described difficulties when seeking care. For instance, fear of stigmatisation and lack of understanding from others regarding the nature of the disease has resulted in some feeling the need to conceal their symptoms and avoid situations where they may have to disclose their diagnosis.38 39 Given the importance of external networks, support-targeted interventions are likely to be beneficial in helping to reduce the stigma and misunderstanding surrounding the condition, thus better supporting people with PH to share their experiences and receive care from others. Moreover, while carers’ needs are all too often overlooked in medical provision, evidence has demonstrated the social and emotional impact PH can also have on carers.40 Therefore, helping to increase carers’ knowledge, confidence and develop helpful strategies of coping is also likely to contribute to further improvements in patients’ HRQoL.
Interestingly, our analysis suggested that greater use of external coping was related to higher levels of health anxiety. While it should be noted that the correlation was low (r=0.19), it is nevertheless consistent with the tendency of people with health anxiety to seek reassurance and support from others as a way of managing anxiety and uncertainty about their symptoms. It is also important to note that this coping approach was the only component that moderated the relationship between HRQoL and both health anxiety and depression suggesting high use reduced the association. While this can be an effective strategy reducing feelings of anxiety in the short term, over-reliance on reassurance-seeking tends to serve an unhelpful maintenance function in the long-term.41 42 However, this needs to be balanced with the need for individuals with PH to be vigilant towards change in their PH symptoms and seek medical care when necessary. It may therefore be important for future research to identify helpful ways for individuals with the disease to receive external support, which does not perpetuate their anxiety.
The final component identified in the analysis represented cognitive and affirmation coping strategies. This consisted of planning, acceptance, self-distraction and positive reframing. One of the most well-known models of coping developed by Lazarus and Folkman9 distinguishes between problem-focused and emotion-focused (also known as active and passive) coping, suggesting problem-focused coping efforts to be an adaptive way of responding to stressors, although this is debated.43 Indeed, qualitative evidence has demonstrated patients with PH engage in extensive planning to help manage the short-term and long-term effects of the disease.44
Importantly, in the present study, cognitive and affirmation coping was found to moderate the relationship between depression and HRQoL, serving a protective function for those who made greater use of this strategy. Helping patients with PH develop cognitive coping skills may therefore be a helpful target for intervention. While there is a dearth of evidence examining psychological interventions in PH, there is some evidence to support their use.23 45 A pilot randomised controlled trial of a self-help cognitive–behavioural therapy (CBT) intervention for anxiety in PH found changes in unhelpful cognitions and behaviours mediated the relationship between intervention group (CBT vs waitlist) and change in anxiety and depression.46 The use of cognitive techniques to alleviate distress is at the core of CBT where positive reframing and distraction are often used, as well as third-wave CBT approaches, such as acceptance and commitment therapy (ACT), which makes use of acceptance and mindfulness. Both types of treatment may be beneficial to patients. Indeed, while there is growing evidence demonstrating a link between psychological factors and patient-reported outcomes, further research is needed. Coping strategies identified in the present sample merits future research to validate proposed components for example using a confirmatory factor analysis. This could also be assessed in specific subgroups of people with PH such as those with diagnosed anxiety or depression, or individuals from non-white communities (120 of 121 self-reported themselves as white in the current study). Longitudinal research would also clarify the relationship between coping styles and challenges associated with PH overtime, for example, experiences of diagnosis, managing the daily demands of the condition and palliative and end-of-life care.
Given that participants self-reported the aetiology of their PH group and WHO function classification, with a considerable proportion reporting to be ‘not sure’ of either, we did not add these as factors in the regression analyses. This approach was taken in order to avoid type 1 and 2 errors. Indeed, PH class has been shown to predict participants’ HRQoL using the emPHasis10 measure.27
A further limitation was the representativeness of our sample. While many of the demographic characteristics were typical of PH samples (eg, a higher proportion of females and those aged 50 and above),47 participants in this study were a self-selected group of relatively highly educated, English-speaking individuals proactively accessing support from a PH organisation online. As such, participants’ involvement in the charity may be impacting on their coping style, as reflected by their scores on seeking emotional and instrumental support.
The findings should be viewed in light of the knowledge that the data were collected in the weeks following the easing of shielding (this involved people in the UK with a pre-existing medical condition that made them more susceptible to COVID-19 (such as PH) spend 12 weeks taking additional infection control precautions). This may have conceivably contributed to increased levels of health-related worry and anxiety within our sample for oneself and for loved ones and therefore a greater need to turn to others for practical and emotional help and support.
This study has provided a unique insight into the coping styles of people with PH. Four distinct coping styles emerged, which were related to measures of depression, health anxiety and HRQoL. While coping by seeking external support as well as cognitive and affirmation coping strategies had a buffering effect on the impact of health anxiety and depression on HRQoL, the tendency to use more passive ways of coping and coping by substance misuse were associated with poorer psychological outcomes. The results are consistent with previous research suggesting people with PH are at a greater risk of experiencing psychopathologies. Findings add to growing evidence base demonstrating the role of psychological factors in predicting quality of life in those with this disease. Understanding the coping styles of this sample has highlighted a range of potential therapeutic targets, supporting the use of psychological treatments aimed at cognitive and emotional processes. Given the observed importance of support systems, patients are also likely to benefit from interventions focused on the provision of information and support to caregivers, although further research is required in these areas.
Data availability statement
Data are available on reasonable request.
Patient consent for publication
This study involves human participants and was approved by Ethical assessment and approval was sought and obtained from the University of Sheffield (035318).
We wish to thank staff at Pulmonary Hypertension Association UK for their assistance in data collection. Also, for the reviewers for raising several important points with an earlier version.
Contributors GHR was the guarantor and responsible for the conception of the study, data collection and analysis, and writing up the study for publication. He approved the final version for publication. ART and NB were responsible for the conception of the study, provided feedback on data collection and analysis and writing up the study for publication. They approved the final version for publication. IA was responsible for the conception of the study, provided feedback on data collection and writing up the study for publication. He approved the final version for publication. BN was responsible for data analysis and writing up the study for publication. She approved the final version for publication.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.