Article Text
PDF
Abstract
Objectives To examine the predictors of medicine-related perceptions towards deprescribing inappropriate medications among older adults in Jordan.
Design A cross-sectional, correlational study.
Setting Data were collected by a graduate nursing student from five outpatient clinics in a selected public hospital in Jordan via inperson interviews 5 days a week over a period of 4 months.
Participants A convenience sample of 200 older adults who regularly visited the outpatient clinics of the selected public hospital for regular check-ups during July 2023 were recruited.
Outcome measures Predictors of patients’ perceived medication concerns, interest in stopping medications, perceived unimportance of medications, and beliefs about medication overuse were examined.
Results Increased perceived medication concerns among patients were significantly associated with older age (p=0.037), lower level of self-rated general health (p=0.002), less perceived care-provider knowledge of medications (p=0.041), higher perceived unimportance of medicines (p=0.018), less collaboration with care providers (p=0.017), being seen by a clinical pharmacist (p<0.001) and an increased number of prescribed medicines (p<0.001). Increased perceived interest in stopping medications was significantly associated with lower levels of self-rated general health (p=0.029), less perceived involvement in decision-making (p=0.013), higher perceived unimportance of medicines (p=0.002), being seen by a clinical pharmacist (p=0.024) and an increased number of prescribed medicines (p=0.001). Furthermore, increased perceived unimportance of medications among patients was significantly associated with more perceived beliefs about medication overuse (p=0.007), more perceived interest in stopping medicines (p=0.001) and greater perceived medication concerns (p=0.001). Moreover, greater perceived beliefs about medication overuse were significantly associated with older age (p=0.018), higher perceived unimportance of medicines (p=0.016), more collaboration with care providers (p=0.038), having post-traumatic disorder (p=0.018) and an increased number of prescribed medicines (p=0.038).
Conclusions The current study examined predictors of medicine-related perceptions towards deprescribing inappropriate medications among older adults. Care providers should discuss the benefits of deprescribing inappropriate medications with their patients to prevent the side effects associated with long-term unnecessary use. Future studies on the effectiveness of an evidence-based deprescribing protocol on minimising the clinical side effects associated with the inappropriate prescription of medications among older adults are recommended.
- Polypharmacy
- Medication Adherence
- Patient Participation
- GERIATRIC MEDICINE
- Decision Making
Data availability statement
Data are available upon 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
Strengths and limitations of this study
The current study used a valid and reliable multidimensional questionnaire, Patient’s Perceptions of Deprescribing (PPoD), among older adult outpatients.
The PPoD has not previously been empirically tested in the Middle East region.
This study did not include open-ended questions, which did not allow for an in-depth assessment and evaluation of the participants’ perceptions.
The study used a cross-sectional design and non-probability sampling method, and the use of a self-reported tool may be associated with reporting bias.
Introduction
According to a recent report by the WHO, the proportion of the ageing population in 2011 was equal to 8% of the world’s overall population.1 Moreover, this number is expected to increase, reaching 16% by 2050.1 In later life, the high prevalence of medical and psychiatric comorbidities among older adults makes symptom management challenging for healthcare professionals.2 3 Since medications are the first choice for managing and preventing chronic illnesses, older adults are more susceptible than other age groups to the inappropriate prescription of medications.4 Older adults living with multiple coexisting chronic conditions are at an increased risk of adverse effects resulting from pharmacological therapy.5 6 Quality use of medications for these individuals involves initiating new medications, modifying dosages and discontinuing ineffective ones.7 Despite the potential benefits of medications in symptom management of chronic illnesses in older adults, there is an increased probability of medication-related hazards among this patient population.8
The inappropriate prescription of medications is a problem of growing interest in geriatrics.9 Inappropriate prescribing involves incorrect prescription, excessive prescription or underprescription of medications.10 It has been reported that approximately 20%–60% of older adults have used at least one medication that could potentially be inappropriate for them.11 The extent of damage caused by the incorrect use of medications is widely recognised and includes adverse impacts on both the individual receiving the medication and the broader healthcare system.12 Furthermore, polypharmacy is prevalent among older adults with comorbid illnesses. Around 20%–65% of older adults take five or more medications per day, increasing the potential for falls, negative drug side effects, hospitalisation and mortality rates.13 Furthermore, the adverse hazards of the inappropriate prescription of medications are associated with increased mortality and morbidity rates, increased risk of falls and cognitive impairment,14 and increased risk of drug-drug interactions and drug toxicity.15
A recent Jordanian study which examined the inappropriate prescription of medications among older adults showed a high prevalence rate of 74.9%.16 Further, another study by the Jordanian Food and Drug Association revealed a prevalence rate of the use of unnecessary drugs equal to 28% among outpatients, with older adults making up 31% of those patients.17 Initiatives have been ongoing worldwide for over 30 years to develop and evaluate interventions targeting the risks associated with the inappropriate prescription of medications, with the focus initially being on discontinuing particular medications deemed problematic for older adults.18 Over time, this has progressed into a comprehensive strategy known as ‘deprescribing’, which assesses medication use by taking into consideration the individual’s additional health conditions, functional ability, treatment objectives and life expectancy.19 20 Deprescribing is one of the approaches adopted to prevent inappropriate prescribing-related pervasive adverse drug events and can be defined as ‘the process of withdrawal of inappropriate medication, supervised by a healthcare professional to improve outcomes and manage polypharmacy’.7 21 One of the potential additional benefits of deprescribing is improving adherence, possibly through reducing polypharmacy and increasing the patient’s engagement in medication therapy management.22 A recent systematic review study reported that deprescribing interventions aimed at decreasing inappropriate polypharmacy may decrease mortality rates, improve patients’ quality of life, decrease adverse drug reactions and promote medication adherence.23
Growing evidence shows that engaging patients in deprescribing decisions increases the likelihood of successful de-escalation.19 23 24 Therefore, understanding patients’ perceptions of their medications and of deprescribing can provide more information about patient-centred healthcare, which is essential to all clinical care.25 Although the primary responsibility for prescribing rests with the clinical provider, understanding the patient’s perspective is critical. Patient behaviours and attitudes regarding medication intake are influenced by numerous factors, including health literacy, socioeconomic status, attitudes, beliefs and the drug’s ability to alleviate symptoms.26 Moreover, understanding patients’ perspectives can allow for a tailored approach to medication deprescribing,27 and attempts to improve clinical procedures for enabling deprescribing have shown greater success when interventions involve the active participation of patients.28 In Jordan, no study to date has examined older adults’ perceptions of deprescribing medications, and therefore, this study aimed to examine the predictors of medicine-related perceptions towards deprescribing inappropriate medications among older adults in Jordan.
Methods
Design
This study had a cross-sectional, correlational design.
Sampling technique and sample size calculation
A convenience sampling method, which involved selecting participants based on their accessibility and willingness to participate, was used in this study. Older adults who regularly visited the outpatient clinics of a selected public hospital in Jordan during July 2023 for regular check-ups and prescription renewal (at least every 3 months) were recruited. The sample size was calculated using a G-power analysis employing an a priori sample size calculator for multiple linear regression. Given an α level of 0.05, an anticipated medium effect size of 0.15, a desired statistical power of 90%, 13 predictors and an additional 15% of the sample size to control for incomplete data, it was calculated that 200 participants would be required to yield significant statistical analysis.
Inclusion and exclusion criteria
The inclusion criteria included being a patient aged 55 years or older, taking at least one medication identified as inappropriate for the patient’s age group or medical condition based on the updated Beers criteria,29 being able to provide informed consent and being able to communicate in Arabic. Meanwhile, patients were excluded from this study if they had: (1) Severe cognitive impairment or dementia that would prevent them from providing reliable information or understanding the study procedures, (2) Severe hearing or visual impairment that would prevent them from communicating with the study team, or (3) A terminal illness.
Ethical considerations
Written informed consent was obtained from the participating older adults prior to data collection, and all collected data were kept private and confidential.
Data collection
Data were collected by a graduate nursing student from five outpatient clinics in a selected public hospital in Jordan via inperson interviews conducted 5 days a week over a period of 4 months. An associate professor in nursing supervised data collection and provided the student with extensive training on administrating the questionnaire, calculating the scores and communicating with older adults. The professor and graduate student checked the reliability of the data collected every 10% of the questionnaire. The sociodemographic (eg, age and gender) and clinical data (eg, comorbidities) of the participants were collected using a demographic data sheet.
The participants’ perceptions of deprescribing inappropriate medications were measured using the Patient’s Perceptions of Deprescribing (PPoD).27 The PPoD consists of 57 multiple-choice questions divided into the following sections: patients’ sociodemographic data, health, medicines, healthcare providers and experience of care provided. For the patients’ medicines, health and healthcare provider sections, patients rated their perceptions using a 5-point Likert Scale ranging from 1=‘strongly disagree’ to 5=‘strongly agree’. The participants’ responses were recorded and analysed to evaluate their perceptions of medication deprescribing. The PPoD consists of eight subscales, including medication concerns (six items), provider knowledge (three items), interest in stopping medications (three items), patient involvement in decision-making (three items), the unimportance of medications (three items), beliefs about medications (four items), trust in a primary care provider (PCP) (five items), and collaboration with PCPs (three items). The total subscale scores were calculated as the average score of the items used to measure each subscale. For this study, only the scores of the subscales related to medication (ie, medication concerns, interest in stopping medications, unimportance of medications and belief about medications) were analysed. The PPoD is a valid and reliable tool, and the Cronbach’s α of the questionnaire subscales in the current study were as follows: concerns (0.94), interest in stopping (0.94), unimportance (0.85) and beliefs (0.94).
The questionnaire was translated into Arabic by two associate professors in English and nursing and then back-translated into English by a professor in linguistics. No significant discrepancies between the two versions were identified, and as for minor differences, the two English versions were reviewed thoroughly until an agreement about the disputed items was reached.
The researchers consulted an associate professor in geriatrics and two professors in clinical pharmacology, who reviewed the translated version of the questionnaire to assess its face and content validity in terms of its equivalent meaning, clarity and cultural relevance. The survey’s content validity index was 0.95, with no recommendations to remove or add any items. All of the consulted professors agreed that all items of the translated questionnaire were clear, relevant, culturally relevant and equivalent in meaning to the original version.
Moreover, a pilot study was conducted prior to data collection to evaluate the feasibility of the study and assess the reliability, feasibility and clarity of the Arabic version of the questionnaire and auditing checklist. The pilot study included 12 older adults, 12% of the desired total sample, and the data collected were not included in the original study. After obtaining their consent, the researcher conducted 15 min face-to-face interviews with the participants in a quiet, private room. The participants were asked to think out loud about the clarity of the questionnaire items, and they reported no difficulty comprehending and responding to the questionnaire items. Similar to the pilot, in the original study data were collected via 15 min face-to-face interviews conducted in a quiet, private room.
Data analysis
Means and SDs were used to describe the continuous sociodemographic variables, and frequency and percentages were used to describe the categorical ones. Multiple-response dichotomy analysis was used to describe the variables measured with more than one option, like comorbidity. The histogram and the Kolmogorov-Smirnov test of statistical normality were used to assess the statistical normality of metric-measured variables. Cronbach’s α internal consistency test was used to assess the reliability of the PPoD questionnaire subscales. Multivariable linear regression analysis was used to examine the significant predictors of the four PPoD subscale scores, namely trust in PCP, PCP knowledge, collaboration with PCP and involvement of patients in decision-making. The association between the independent predictor variables and their dependent outcome variables was expressed as unstandardised β coefficients with their associated 95% CIs. Outlier analysis with residual analysis, Mahanalobis distance, and β indices were used to identify any outlier case (n=1) excluded from the regression analysis models. A licensed copy of the SPSS IBM statistical computing software V.28 was used for the statistical data analysis, with an α significance level of 0.050.
Patient and public involvement
Patient and public involvement (PPI) representatives (Wafa Ajlouni, Maha Ali and Doha Hijazi) were regularly involved in the study design and interpretation of the results in multiple meetings. WA and DH were actively involved in developing the study materials. Prior to data collection, a meeting was held with the PPI representative (DH) and a local pharmacist (MA) to ensure that the study materials were appropriate for older adult outpatients. A pilot interview was conducted with the PPI representative to ensure the clarity and appropriateness of the study questionnaire. Regular meetings were held between the researchers (MJR and AMA) and a PPI representative (WA) to discuss the analysis plan and interpret the findings. The researchers (MJR and AMA) and the PPI representative (DH) presented the results and led group discussions about interpreting and disseminating the results.
Results
Descriptive analysis of patients’ sociodemographic characteristics
Two hundred and five participants were recruited in the study. Only five participants were excluded from the analyses due to them not completing the questionnaire. Two hundred older adults completed and returned the study questionnaire and were therefore included in the analysis. Table 1 displays the patients’ sociodemographic and clinical characteristics. Most of the participating older adults were male (55%), married (65%) and educated to primary school level at least. Most of the participants had several medical and psychiatric comorbid problems, such as anxiety disorder (36%), diabetes mellitus (75%) and hypertension (84%).
Descriptive analysis of patients’ sociodemographic characteristics (n=200)
Predictors of patients’ perceived medication concerns
Table 2 displays the multivariable linear regression analysis of the patients’ perceived medication concerns and their associated factors. As shown in table 2, a statistically significant regression equation was found (p<0.0001). Almost half of the variables entered into the regression model were found to be statistically significant predictors (p<0.05). For example, self-rated level of general health had a significant negative association with mean perceived medication concerns, β coefficient =−0.144, p=0.002, indicating that a greater sense of general health predicted less perceived medication concerns.
Multivariable linear regression analysis of patients’ perceived medication concerns
Predictors of patients’ perceived interest in stopping inappropriate medications
Table 3 displays the multivariable linear regression analysis of the patients’ perceived interest in stopping inappropriate medications and the associated factors. As shown in table 3, a statistically significant regression equation was found (p<0.0001). Five variables entered into the regression model were found to be statistically significant predictors (p<0.05). For example, patient’s perceived involvement in medication prescribing/deprescribing decision-making had a significant negative association with perceived interest in stopping medications, β coefficient=−0.118, p=0.013. Therefore, it may be inferred that greater involvement of the patients in the decision-making process related to their medications prescribing/deprescribing predicted less perceived interest in stopping inappropriate medications.
Multivariable linear regression analysis of patients’ perceived interest in stopping inappropriate medications
Predictors of patients’ perceived unimportance of medications
Table 4 displays the multivariable linear regression analysis of the patients’ perceived unimportance of medications and the associated factors. As shown in table 4, a statistically significant regression equation was found (p<0.0001). Only three variables entered into the regression model were found to be statistically significant predictors (p<0.05). For example, perceived medication overuse was associated positively and significantly with perceived unimportance of medications, β coefficient=0.138, p=0.007. This indicated that higher levels of perceived medication overuse predicted significantly higher levels of perceived medication unimportance. Moreover, the analysis model suggested that the patients’ perceived interest in stopping medicines was associated positively and significantly with their perceived unimportance of medications score, β coefficient=0.196, p=0.001. This indicated that greater interest in stopping medicines predicted significantly higher levels of perceived unimportance of medications.
Multivariable linear regression analysis of patients’ perceived unimportance of medications
Predictors of patients’ perceived beliefs about medication overuse
Table 5 displays the multivariable linear regression analysis of the patients’ perceived medication overuse and the associated factors. A significant regression equation was found (p<0.0001). Most of the variables entered into the regression model were found to be statistically significant predictors (p<0.05). For example, patients’ perceived unimportance of medicines had a significant positive association with their mean perceived medicine overuse, β coefficient=0.237. More specifically, as the mean perceived unimportance of medicine score rose by 1 point on the Likert Scale, mean perceived medicine overuse increased by 0.237 points, p=0.016.
Multivariable linear regression analysis of patients’ perceived beliefs about medication overuse
Discussion
This study is one of very few to use the PPoD tool to measure older adults' perceptions of deprescribing medications, rendering it difficult to compare the current study results to those of previous studies. However, despite the use of different research tools and methodologies, the study results may be discussed in light of previous studies related to the topic of the current study.
Factors associated with patients’ perceived medication concerns
The findings of the current study revealed that patients’ age, PCP knowledge of medications, perceived unimportance of medicines, collaboration with PCP, self-rated general health, number of prescribed medications and hospital admission rate are statistically associated with patients’ perceived medication concerns. Patient age was found to play a significant role in influencing patients’ medication concerns, as different age groups may have varying health needs, perceptions and attitudes towards medications.30–33 For instance, older adults tend to be more concerned because they may use multiple medications or have age-related health issues.16 34 35
Moreover, the findings showed that PCP medication knowledge is associated with patients’ medication concerns, as patients may be more confident and less concerned about their medications when they perceive their PCPs to be educated and skilful. As for perceived medicine unimportance, our findings suggested that the patients' perception of the importance of their medications is negatively associated with their overall medication concerns. Therefore, the less important that patients perceive their medications to be, the more concerns or reservations they are likely to have about them.
Self-rated general health, a subjective measure, was also found to be connected to medication concerns, with patients who perceived their overall health to be poor being more likely to be concerned about their medications. This may be attributed to the fact that medications significantly impact health, and so patients may be concerned about their poor overall health deteriorating further. Another factor found to significantly influence medication concerns is the number of medicines taken. Specifically, a significant positive association was identified, indicating that patients with a greater number of prescribed medications may express more concerns related to drug interactions, potential hazards and complex medication management. Moreover, the findings suggest that patients with a history of recurrent hospital admissions may express more concerns about their medications, which may be attributed to their past experiences with medication adverse events and the healthcare system or concerns about treatment effectiveness.
In summary, patients’ concerns about medication were found to be significantly associated with multiple factors, including sociodemographic, PCP-related and medicine-related factors, and self-rated general health. Alleviating medication-related anxiety and improving patients’ experiences and satisfaction could be achieved by targeting these factors through individualised patient-centred care, effective communication and patient education. Despite the limited available research using the PPoD tool, the current findings were supported by the findings of previous studies. For example, Smaje et al 36 found that PCP knowledge and the number of medications were statistically associated with older adults’ deprescribing medications concerns. Moreover, Robinson et al 24 found lack of knowledge, fear of adverse events, collaborative approaches and the number of medications taken to be statistically associated with older adults’ medication concerns.
Factors associated with patients’ perceived interest in stopping medications
The current study findings revealed that patients’ involvement in the decision-making process, perceived unimportance of medicines, medication concerns, self-rated general health and number of prescribed medications are significantly associated with their interest in stopping medications. It can be inferred from the findings that older adults who actively participate in choosing their prescriptions may be more willing to stop a particular medication. Further, the less important that older adults perceive a particular medication to be, the more interested they may be in stopping it. Furthermore, the finding that older adults with higher levels of medication concerns were more willing to stop these medications may be attributed to the perceived risks or fear of adverse effects. Moreover, older adults who believe their health status to be poor may lean towards stopping some prescriptions, either to regain control over their health or because they believe the medications to not be as important as they had thought. Finally, the current findings suggested that older adults with a greater number of prescribed medications have a greater interest in stopping them.
Many previous studies have explored older adults’ attitudes towards stopping their medications, with some of our findings supported by these studies. Similar to the current study, previous studies37–39 have shown that the number of medications taken by older adults and their concerns about medications are significantly associated with their willingness to stop medications. Furthermore, Oktora et al 31 found that older adults’ willingness to stop their medications was significantly associated with their medication concerns and perceived burden of medications, which is in line with the present study findings. Meanwhile, Oktora et al 31 also reported older adults’ educational level to be associated with their interest in stopping medications,31 which is not consistent with the current study findings. However, it is noteworthy that the study of Oktora et al 31 was conducted in Indonesia and used a different tool, the Revised Patients' Attitudes Towards Deprescribing (rPATD), which may explain the inconsistency in the findings.
Contrary to Oktora et al,31 Pereira et al 33 used the rPATD and found older adults’ age, gender and medication concerns to be significant predictors of their willingness to stop medications. Similarly, Kua et al 40 and Crutzen et al 41 found patients’ medication concerns, gender and educational levels to be significantly associated with their willingness to stop medications. In a similar vein, Shrestha et al 42 reported patients’ age and medication concerns to be statistically associated with their interest in stopping medications. It is noteworthy that the inconsistency between previous findings and ours pertaining to factors influencing patients’ perceived interest in stopping medications may be attributed to the use of different measurement tools, different research methodologies or different settings.
Factors associated with patients’ perceived unimportance of medications
The current study found beliefs about medication overuse, interest in stopping medications and medication concerns to be significantly associated with patients’ perceived unimportance of medications. This finding indicates that patients’ beliefs about medication overuse are essential in forming their perceptions of medication unimportance, which is consistent with previous research.35 43 The finding that the patients’ desire to stop medication use contributed to their overall perception of medication unimportance could be associated with their concerns about the efficacy, side effects or factors related to stopping medication. These findings can be justified in light of various previous research studies.14 35 44 For example, Davies et al 14 found that most older adults prefer to stop their medications if they perceive them to be inappropriate or unimportant. The current study findings also suggested that identifying specific medication concerns may aid PCPs in tailoring interventions to improve patient adherence and confidence in medication efficacy, which is supported by previous research.17 43 44 For example, Nusair et al 17 reported that most older adult patients in Jordan perceived medications with significant potential drug-drug interactions as unimportant medications that should be stopped.
Factors associated with patients’ beliefs about medication overuse
The current study found patients’ age, perceived unimportance of medications, post-traumatic stress disorder and the number of prescribed medications to be significantly associated with perceived medication overuse. Older adults may have different attitudes towards medications as compared with younger adults42 and may perceive and interact with healthcare, prescription and deprescription in different ways. Age-related differences in life experiences, cultural background and health literacy can all influence patients’ perceptions of the benefits and possible risks of medication.31 43 Furthermore, these findings suggest that understanding patients’ adherence and use patterns requires an awareness of their perceptions of the value or necessity of medications. Patients may be more likely to overuse or underuse certain medications if they believe them to be unimportant.31 43 Thus, encouraging optimal usage may depend primarily on addressing the needs of patients and communicating the need for prescribed medications. Medication adherence is known to be impacted by the intricacy of prescription regimens, and multiple prescription medication regimens can make it difficult for patients to follow their treatment goals and hence cause misconceptions about the abuse or underuse of medications.31 42 Thus, when prescribing several medications, PCPs should monitor patients frequently, simplify prescription regimens when possible and give clear instructions.
Study limitations
There were several limitations to the present study. First, it used a cross-sectional design and a non-probability sampling method, and the use of a self-reported tool may be associated with reporting bias. To expand the generalisability of the findings, replicating this cross-sectional study with a larger sample and different inpatient categories is recommended. Further, an experimental study examining the effectiveness of an evidence-based deprescribing protocol in minimising the clinical side effects associated with inappropriate medication prescriptions among older adults is also recommended.
Conclusions
Inappropriate prescription of medications is one of the most common polypharmacy-related issues among older adults. The current study explored older adults’ perceptions towards deprescribing inappropriate medications, with the findings indicating a high prevalence of polypharmacy among older adults in Jordan. Moreover, the study found that older adults’ perceived interest in stopping medications, beliefs about medicine overuse, perceived unimportance of medicines and medication concerns are significantly influenced by a variety of sociodemographic (eg, age), clinical (eg, comorbidities) and PCP-related factors (eg, knowledge of medications). Therefore, PCPs should discuss the benefits of deprescribing inappropriate medications with their patients to prevent the side effects associated with long-term unnecessary use. The long-term use of inappropriate medications by older adults should be carefully evaluated. Future studies on the effectiveness of an evidence-based deprescribing protocol, such as The Track and Trigger Treatment Protocol18 on minimising the clinical side effects associated with the inappropriate prescription of medications among older adults, are recommended.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the Institutional Review Board at Jordan University of Science and Technology (approval number 378-2023). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
The authors thank everyone who assisted with participant recruitment and all the individuals who participated in the study, without whose contributions this study would not have been feasible.
References
Footnotes
Contributors MJR and AAG contributed to the conception and study design, prepared the study protocol, drafted the manuscript, performed the literature review, carried out statistical analyses, interpreted the results, and acquired and managed the data. MJR, AAG and AAH wrote the manuscript and reviewed the manuscript for critical revisions. All authors approved the final manuscript. MJR is the guarantor.
Funding This research was funded by the Deanship of Research at Jordan University of Science and Technology (grant number 20230435).
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.