Article Text
PDF
Abstract
Introduction Multiple punctures during peripheral venous catheter (PVC) placement increase the risk of complications. Scoring for adult difficult intravenous access (A-DIVA Scale) exists but has never been assessed in the framework of a care algorithm (scoring associated with a new decision-making tree for puncture conditions, the A-DIVA Tree). We seek to implement an catheter placement algorithm to decrease the mean number of punctures per patient. The algorithm will be adjusted based on obstacles and levers revealed by the analysis of clinical data. The benefits of the algorithm will be assessed using a step-by-step implementation of the approach.
Methods and analysis 794 PVC placements will be recorded in two inclusion centres (50%/50%). In phase I, 297 PVC placements will be collected, and 16 individual semistructured interviews will be conducted to evaluate the centres’ practices. In phase II, 200 PVC placements will be recorded to assess the impact of the A-DIVA Scale alone. The interphase will allow preliminary results based development of the A-DIVA Tree. In phase III, 297 PVC placements will be recorded to assess the impact of the algorithm on the mean number of punctures per patient.
Ethics and dissemination The study and related consent forms were approved by an institutional review board (Comité de Protection des Personnes Sud-Méditerranée I) on 25 April 2023 under reference number 2023-A00223-42. The results will be disseminated in the form of original articles, presentations and guidelines.
Trial registration number NCT05935228.
- ANAESTHETICS
- ULTRASONOGRAPHY
- QUALITATIVE RESEARCH
- Primary Health Care
- ACCIDENT & EMERGENCY MEDICINE
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
A mixed (quantitative and qualitative), multicentre, prospective study to develop and assess a peripheral venous catheter (PVC) placement algorithm.
A step-by-step process with control and experimental groups to analyse the practices, constraints and expectations of practitioners regarding PVC placement.
A 30-month study involving a large number of healthcare professionals.
An analysis based on semistructured individual interviews, data collection on care procedures, sociodemographic data and multiprofessional workgroups.
A multiphase study in which each phase must be blinded to subsequent ones to avoid spontaneous modifications in healthcare professionals’ practice.
Introduction
Two billion peripheral venous catheters (PVCs) are used per year worldwide.1 Nearly 80% of hospitalised patients have at least one PVC placed in the hospital setting, making catheterisation the most frequently performed invasive procedure.2 Nearly a quarter of all people requiring vascular access have difficult intravenous access (DIVA).2 3 Multiple complications, such as obesity, intravenous drug addiction, vascular pathologies and cancers, are a source of difficulty and failure in vascular access placement and may result in the need for alternative puncture methods.3–7 The consequences on an individual and collective scale can be significant, with delays in care or prolonged hospital stays. For example, patients require a PVC showed systemic infections (rate: 0.5/10008), infiltrations and venous thrombosis (23% and 12%9).1 Additionally, multiple iterative punctures increase the risk for complications, such as haemorrhages and haematomas, local or systemic infections, skin or nerve damage, pain and psychological trauma and gas embolism.1–3 8–16 Complications associated with DIVA affect practitioners’ workflow, particularly that of nursing staff responsible for providing this care. The feeling of incompetence inherent in failures and complications is also a source of discouragement for practitioners.17 Finally, in the current context of hospital tension, these daily difficulties can contribute to the increased pressures that healthcare institutions encounter in terms of staffing turnover and economic level.18
Because of the frequency of DIVA and its related complications, substantial interest in this topic has arisen in the scientific community over the last decade, and the number of publications related to the subject has significantly increased.19 Van Loon et al developed and tested a scale called the A-DIVA Scale, for adults, to assess the risk of DIVAs in patients in 2016 (modified in 2019).15 16 The scale comprises five items (visibility of the vein, palpability of the vein, diameter of the vein, a history of DIVAs and the practitioner’s expectation of difficult intravenous cannulation) and allows patients to be classified according to the risk of failed puncture (online supplemental table 1). The number of failures has been found to increase with the A-DIVA Score, suggesting the effectiveness of this scale in anticipating practitioner difficulties.16 20–22
Supplemental material
All the aforementioned studies mention the risks inherent in the placement and use of PVCs and the need to assess people likely to experience DIVAs along with presenting alternative solutions.23–25 However, these actions occur after failures in an unanticipated and non-consensual manner. The techniques and tools developed thus far only offer a partial solution to the problem of multiple punctures that they are intended to prevent. Before any evaluation or performance of this invasive procedure, it is imperative to develop an organisation for managing peripheral vascular accesses based on the individuality of each patient and the reality of the care sites, particularly for patients with DIVAs. This need could be met by developing a gradual and specific response to the issue of multiple punctures. Therefore, this project aimed to create and evaluate a specific algorithm consisting of a risk assessment (the A-DIVA Scale) and a co-construction decision-making tree (the A-DIVA Tree). The latter will consider the realities and possibilities of practitioners to respond to the need for risk prevention by anticipating the actions to be taken and avoiding bad decisions and disorganisation (an example of the A DIVA Tree is provided in online supplemental figure 1). This decision-making tree would permit systematically (and without loss of time) the choice of the correct operator (with sufficient seniority and technical background) and a good technique linked to the individual patient assessment. With it, the risk of associated care complications (such as infections, haematomas, pain, psycho traumas, etc) would decrease, as would care time, human and economic costs.
Supplemental material
This original algorithm would offer a complete system, improving patient safety by reducing the number of punctures required for PVC placement and their exposure to the associated risks.
Objectives, outcomes and hypothesis
Primary objective
The primary objective is to assess the effect of implementing a PVC placement algorithm versus standard care on the mean number of punctures per patient.
Primary outcome
Number of punctures performed per patient after implementation of the algorithm.
Primary hypothesis
Implementing and systematically using a management algorithm for patients requiring peripheral venous access will reduce the number of punctures needed per patient to achieve PVC placement.
Secondary objectives
To co-construct a catheter placement algorithm based on the venous access difficulty score.
To assess the acceptability and feasibility of using the A-DIVA Score in isolation and the score in conjunction with the catheter placement algorithm.
To describe how patients feel in terms of pain during care and overall comfort before and after the implementation of a catheter placement algorithm.
To evaluate the effect of implementing the venous access difficulty score on the average number of venipunctures per patient.
To assess operators’ professional practices in terms of insertion technique.
To assess operators’ professional practices in terms of number of operators required for catheter insertion.
To assess operators’ professionals practices in terms of frequency of second-line referral to another practitioner.
To assess operators’ professional practices in terms of success or failure.
To evaluate the strategies implemented in the event of difficulties encountered when inserting PVC.
Secondary outcomes
Number of participants who approve the co-constructed algorithm.
To meet this objective, caregivers’ appreciation, use and feelings regarding the vascular access difficulties score and algorithm will be measured using a 23-item questionnaire scored on a 4-point Likert scale (‘strongly disagree’, ‘disagree’, ‘agree’, ‘strongly agree’). This questionnaire will also measure the effect of the vascular access difficulties score and the algorithm on practices.
Patient pain score measured immediately after vascular puncture (phase of potential acute pain during catheter insertion) using a self-assessment tool (numerical scale ranging from 0: no pain; 10: maximum pain).
Number of venipunctures performed per patient after implementation of the venous access difficulty score.
Insertion technique used for catheter placement.
Number of operators required for catheter insertion.
Frequency of second-line referral ton another practitioner.
Number of successful/failed punctures.
Strategies implemented in the event of difficulties will be measured using a multimodality question (several answers possible): call another practitioner, use an alternative insertion technique, choose another type of venous line and give up.
Methods and analysis
Clinical investigation design
This is a mixed (quantitative and qualitative), multicentre, non-randomised, quasi-experimental, before-and-after, open, cross-sectional study with control and experimental groups comprising three principal phases. The project’s expected duration is 30 months, starting in September 2023 and ending in March 2026 (projected).
Phase I, the observational phase (4 months), will provide information on the actual practices (including the mean number of punctures for successful placement) and organisation. The main objective of this phase is to provide an in-depth and detailed description of existing practices for PVC placement in the centres using quantitative and qualitative data. Phase I relies on the principles of a feasibility study since the aim is to investigate the context to adapt the implementation of the intervention in development to the key context dimensions of healthcare and the health system, as recently recommended by the medical research council.26 This phase will consist of a systematic collection of PVC placement practices to obtain quantitative data (point-of-care questionnaires) and individual semistructured clarifying interviews (until 16 interviews, 8 per centre) with professionals. The aim of the clarifying interview method is to encourage the verbal description of how a task is effectively realised. Verbalising is rarely descriptive, mostly refers to generalities, and almost never refers to how an action is realised. The principles of the clarifying interview method consist of focusing the participant on a single real situation (time and space situated) and directly encouraging the verbal elaboration of procedural action.27 To apply a maximum variation sampling strategy, we identified the healthcare professionals to include in the semistructured interviews as the main dimension of variation. Thus, we included nurses, specialised nurses (anaesthetic nurses) and physicians attached to the department will be the nurses, specialised nurses (anaesthetic nurses) and physicians attached to the department.28 The semistructured individual interview guide for professionals is shown in online supplemental figure 2.
Supplemental material
Phase II, an interventional phase (4 months), aims to evaluate the impact of implementing the A-DIVA Scale alone (classification score for venous access difficulties) on practitioners’ approaches to peripheral venous access placement. It consists of a systematic collection of peripheral venous access placement practices to obtain quantitative data (including the mean number of punctures for successful placement) and an ad hoc open-ended questionnaire on what is known, believed and achieved with the A DIVA Scale.29 30 The main objective of this phase is to recognise the potential appearance of changes in practices through implementing a systematic assessment of vascular access difficulties alone and optimising its implementation in the future.
The interphase (between phases II and III: 8 months) will allow us to analyse the data collected in phases I and II to develop the A-DIVA Tree. The scientific team (principal investigator and the Centre d’Investigation Clinique-Epidémiologie Clinique) will analyse the data. Then, two workgroups (one group with varied profiles of three to six practitioners per centre and the corresponding principal investigator) will co-construct the algorithm. Data will be submitted to the working group to inform the construction of the algorithm. In particular, discussions should enable us to define the organisation of PVC insertion in terms of care stratification, material and service resources. As operator’s skills and seniority have a tangible impact on the success in placing PVC, this phase will also highlight the operator’s characteristics needed to achieve the objective of reducing the average number of punctures per patient. The deliberations of the two working groups will be compared and reconciled and resubmitted to the two groups until complete validation.
Phase III, which is interventional (4 months), will allow the implementation and assessment of the effectiveness of the complete algorithm (A-DIVA Scale and A-DIVA Tree). It will consist of a systematic collection of PVC practices to obtain quantitative data (including mean number of punctures for successful placement), an ad hoc open-ended questionnaire and a satisfaction questionnaire (23 questions, Likert scales) addressed to professionals distributed at the end of phase III (see online supplemental table 2). The main objective of this phase is to evaluate the impact of the algorithm on the efficiency of PVC placement. The criterion for objectively determining the impact of the algorithm is the increase or decrease in the mean number of punctures required for the successful placement of the PVC.
Supplemental material
The postprocessing phase (10 months) will allow the analysis of all the data (from phase I to phase III) to identify preliminary indicators of the process from the perspective of the deployment of the algorithm in the framework of a national implementation trial, the production of reports and scientific articles and the drafting of the final report of the algorithm for peripheral venous catheter placement (ALCOV) project.
The detailed study schedule is shown in figure 1.
Description of estimated timeline. A-DIVA, adult difficult intravenous access.
Investigators and practitioners
Two principal investigators will co-ordinate and lead the study (one per centre). They will not directly participate in the quantitative data collection.
All the local practitioners (nurses and physicians) will contribute to the data collection when performing a puncture for peripheral venous access (participant enrolment and point-of-care questionnaire completion).
Recruitment of patients
All eligible patients will be recruited after receiving information about the study and giving oral and written consent (online supplemental figure 3). The inclusion and exclusion criteria are presented in table 1.
Supplemental material
Patients’ inclusion and exclusion criteria for the study
Recruitment will take place at the University Hospital of Nancy (CHRU de Nancy) and Pont-à-Mousson Hospital. Given the nature of the study (based on the collection of the care procedure: placement of PVCs), some patients may be enrolled several times during the data collection period. They will be given a new anonymisation number for each procedure, and each inclusion will be considered as a single event.
We will include 794 patients over 12 months, made possible by the high potential for inclusion at the selected sites. For example, the operating theatre of CHRU de Nancy performs care activities requiring the placement of 100–150 PVCs per week (5000–7800 placements per year). At the Pont-à-Mousson emergency department, in 2018, 881 PVCs were placed within 3 months (over 3500 placements per year).
Nature of the collected data
The practitioners who perform the care will fill in the point-of-care questionnaire, allowing the collection of quantitative data related to the placement of PVCs. This dataset will provide an understanding of the functioning and issues of this care (phase I control group) and the impact of the introduction of tools (phase II, the A-DIVA Scale alone, and phase III, the A-DIVA Scale and the A-DIVA Tree) on the number of punctures required for the successful placement of PVCs. The content of the information given to practitioners will be adapted to minimise any impact on their behaviour during the punctures and any bias in declaration and interventions. Details of the collected quantitative data are provided in table 2.
Description of quantitative data collected throughout the study
A human and social sciences research engineer will conduct semistructured interviews to collect practitioners’ perspectives. The exchanges will be audio recorded, anonymously transcribed and stored on a secure server, accessible only by research team members (table 3).
Description of qualitative data collected throughout the study
Clinical procedure
Phases I–III
Patients who agree to participate in the study will receive, depending on the phase, PVC placement according to the protocol of the institution (ie, phase I) or an assessment of difficulties of vascular access risk (DIVA Scale) and PVC placement according to the protocol of the institution (ie, phase II) or a PVC placement according to the algorithm (A-DIVA Scale scoring and DIVA tool based decision for puncture, that is, phase III). After PVC placement, depending on the phase, the practitioner will complete the point-of-care questionnaire.
All the PVC placements during the study period will be annotated as to whether the patient was included in the study to evaluate the sample’s representativeness. Each workstation will be equipped with a document allowing the rapid tracing of all PVC placements. The procedure for collecting data from patients is shown in figure 2.
Clinical procedure flowchart. A-DIVA, adult difficult intravenous access.
Participation durations
Patients
The duration of participation for each patient corresponds to the necessary time to insert a PVC, which is between 5 and 60 min.
Practitioners
The duration of practitioners are as follows:
The mean duration of an individual semistructured interview is approximately 60 min.
Participation in a workgroup will also take 60 min.
Participation in satisfaction questionnaires at the end of the process is estimated to take 2–5 min.
The characteristics of the study phases are summarised in online supplemental table 3.
Supplemental material
Statistical design
Sample size justification
Van Loon et al found a mean number of punctures per patient of 1.3.16 With an alpha risk of 5%, a power of 80%, an SD of 0.85 and a two-sided formulation of the hypotheses, 564 patients are needed to show a 15% decrease in the mean number of punctures per patient following the implementation of the algorithm (we believe that a 15% decrease would be clinically relevant). Assuming that 5% of patients could be excluded, 594 patients should be included. The two inclusion centres will divide these patients equally (50/50). This number of subjects was calculated to meet the primary objective (comparison of the mean number of punctures per patient in phase I to phase III).
To be able to carry out phase II and meet specific descriptive secondary objectives (description of the effect of the implementation of the venous access difficulty scale on the mean number of punctures per patient), 200 other patients (100 per centre) will be included for the phase II study. This number of patients will allow sufficiently accurate estimates of the measured data.
Overall, 794 patients will be included in the study.
In addition, health practitioners will be included to conduct 16 individual semistructured interviews. Data saturation is achieved when concepts and subconcepts cannot be further specified with additional data. Data collection in the context of inductive thematic analyses occurs until saturation point.31 We chose to estimate sample size based on literature without assessing and reporting thematic saturation using a calculation approach such as the one proposed by Guest et al.32 A number of interviews ranging from 11 to 20 corresponds to a satisfactory sample size for qualitative studies to obtain satisfactory findings.33–35 We reached thematic saturation with the inclusion of 13 participants. With narrowly defined objectives, as is the case in our study, saturation can be achieved with a relatively small sample size.
Statistical analyses
Preliminarily, we will verify variable distributions, assess outliers, check for errors in coding and check for missing data. Characteristics of the study sample and outcomes (patient and centre characteristics, puncture data, pain and comfort score) will be described as the mean (SD) or median (Q1; Q3) for quantitative variables (according to their distribution) and the number (%) for qualitative variables. Characteristics of the sample and outcomes will be reported per phase. Characteristics of the sample will be compared between phases using analysis of variance or the Kruskal-Wallis test for quantitative variables and χ² or Fisher’s exact test for qualitative variables.
Then, the effect of the implementation of the algorithm (phase III vs I) and of the A-DIVA Scale (phase II vs I) on the mean number of punctures per patient will be evaluated using multilevel linear regression models (levels=centres and practitioners). First, crude models will be conducted. Then, models will be adjusted for potential confounding factors (sociodemographic and/or clinical characteristics of patients). The analyses will be conducted on an intention-to-treat basis. In case of missing data, the maximal bias hypothesis will be assumed. No intermediate analysis will be planned. If deviations from the algorithm are observed, a per-protocol analysis will be performed, considering only patients for whom there was no deviation.
The data from the interviews will be audio recorded and integrated. We will follow a general inductive approach to proceed to a descriptive thematic analysis.36 Thus, we will identify themes from the participant discourse to properly capture their perception of the obstacles and levers during the implementation of the algorithm.37 Data will be double-encoded to ensure the reliability of the qualitative analysis. Disagreement between the two coders will be resolved by discussion to reach a kappa value from at least 0.81 per category. Data analysis will be performed using NVivo V.11. The same overall methodology orientation for the thematic analysis process will be applied to study the answers to the open-ended questions in phases II and III.
These data will be compared according to phase (I, II or III) using an expected level of statistical significance at 0.05. The analyses will be conducted in R.
Ethics and dissemination
Ethical permission
Personal data related to the study’s participants will be processed.
These data will be collected and processed solely on the legal basis provided for by the regulations in the context of the performance of the CHRU de Nancy’s public interest missions, in particular those related to ensuring and contributing to research and innovation (Article 6.1 in paragraph of the RGPD).
Per Article L.1122-1 of the CSP, each person will receive information before receiving any specific experimental procedure. This information will be summarised in a written document to the person whose consent is sought. The patient’s consent will be obtained orally and in writing, dated and signed.
In accordance with Article L.1123-6 of the Public Health Code, the research protocol has been submitted by the promotor to the Institutional Review Board (Comité de Protection des Personnes (CPP)). The study and related consent forms were approved by CPP Sud-Méditerranée I on 25 April 2023 with reference number 2023-A00223-42. The study has also been registered in the clinicaltrials.gov registry under the trial number NCT05935228.
Protocol amendment
The sponsor will send requests for substantial modifications to the relevant regulatory authorities for authorisation and/or advice.
Data curation
Personal data relating to the participants will be stored in the information systems of the sponsor, the investigating centre or the practitioner involved in the investigation until the product studied is placed on the market or 2 years after the last publication of the results of the research or, in the absence of publication, until the final report of the study is signed. They will then be archived on paper or on a computer for the period of time required by the regulations in force.
Dissemination
This study will assess the relevance of using an ALCOV. First, these results will be disseminated in the form of an original article. Subsequently, they will enable the design of a national clinical implementation trial assessing the impact of the algorithm on the number of punctures (especially for patients with difficult venous accesses). Finally, this work could provide some guidelines to improve the security and quality of care.
Data monitoring
The level of monitoring is adapted and weighted according to the estimated risk for the person undergoing the research and the logistical complexity, resources and impact of the research results. The monitoring visits (implementation, follow-up and closure) will be conducted by the Research and Innovation Department of CHRU de Nancy (or a representative of the sponsor mandated for this purpose). They will be conducted in accordance with the monitoring plan specific to the research and adapted to the level of monitoring previously defined, as well as the standard operating procedures. Depending on the monitoring reports and the deviations observed, the sponsor reserves the right to modify the planned monitoring level.
Risk and benefits assessment
Potential risk
Potential risk is placing PVCs is an invasive procedure, which leads to skin and vascular injuries. The risks inherent in this treatment are numerous, with more or less severe and more or less frequent impacts, for example, discomfort/pain, skin/nerve/vascular lesions, oedema, haemorrhages, infections, haematomas, psychological trauma, fear/anguish and gas embolism.1 9 10 13 14 38 39 Most of these consequences are minimal and are resolved almost instantaneously in a few minutes to a few days. In sporadic cases, complications may arise that require management that was not initially planned. The individual and collective (family, caregiver, economic) impacts could then be significant. No risk inherent in the puncture should be ascribed to the study because the patients included in the study are to receive a PVC as part of their care. Alternative PVC placement techniques are commonly used as part of routine care.
Anticipated clinical benefits
Benefits for patients
The implementation of the algorithm should optimise patient care with an expected decrease in the duration of care, a reduction in pain and discomfort (by reducing the number of punctures and adding a pain and comfort scale) and a reduction in risks (haemorrhagic, infectious, delays in care, psychological trauma).
Benefits for practitioners
A valuation of skills and experiences and a reduction in exposure to the risk of failure through the use of adapted tools should be the source of increased confidence and competence for practitioners.
Institutional benefits
The puncture-related costs will decrease due to reduced material consumption, a decrease in the number of undesirable events, workflow improvement and a reduction in the time dedicated to the puncture.
Benefit–risk ratio
The benefit–risk ratio analysis indicates that this study is of significant interest. It presents few additional inconveniences (lengthening the procedure for completing the point-of-care questionnaire by a few minutes) on top of the standard procedure. Nevertheless, it proposes improvements likely to reduce the individual and collective risks incurred.
Final research report
The principal investigator and the mandated biostatistician will collaboratively write the final research report. This report will be submitted to each of the investigators for review. Once a consensus has been reached, the final version will be endorsed with the signature of each of the investigators and sent to the promoter as early as possible after the effective end of the research.
Patient and public involvement
We did not involve patients or the public in the design, conduct or reporting of our research.
Discussion
This study aims to develop and evaluate an algorithm for decreasing the mean number of punctures per patient during the placement of PVCs. The collection of preliminary indicators from the ALCOV Study will also be used to optimise the parameters of a future national implementation trial in care services.
The first limitation of this study is that the participant practitioners will be solicited over a long period (three data collection phases of 4 months each and working groups during the interphase). The duration is sample size adjusted but may lead to a decrease in their interest in the study. Regular requests and progress reports are planned to prevent the participant practitioner’s disinterest. Displays will be used as anchor points to facilitate the use and understanding of the study tools. The practitioners’ feelings will be collected throughout the phases and will serve as a barometer to adjust the necessary follow-ups. Last, the variety of requests (workgroups, individual semistructured interviews) and the involvement of practitioners in decision-making will promote mobilisation.
The second possible limitation of this study is related to the practitioners’ knowledge of the purpose of the study. Indeed, there is a risk that they will anticipate and thus modify their practices, making the data and interpretations obsolete. To address this difficulty, each stage will be blind to the next, with practitioners only being aware of the details of the phase in which they will be involved. Thus, their understanding of the project will be built up step by step until a global vision is achieved in phase III.
This project aims to develop a prevention-based approach to optimise PVC placement. Catheter placement is the most common invasive intervention in hospitals, and the results may lead to a major improvement in the quality and safety of PVC placement.
Ethics statements
Patient consent for publication
Acknowledgments
The authors would like to thank Dr Hossu for her help at the beginning of the project, Mr Fantin for his contribution to the linguistic proofreading and Nature's task force for the final proofreading.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @MannevilleF
Contributors CS: Creation and development of the original idea, bibliography, methodological and scientific analysis, drafting and correction of the protocol and study tools; FM: quantitative and statistical methodological analysis and correction of the protocol; LR: qualitative methodological analysis and correction of the protocol; AO: drafting and development of ethical and regulatory aspects and proofreading of the protocol; SK: development, correction and drafting of the protocol and development of partnerships; HB: construction of the project, development of partnerships, scientific analysis and proofreading of the protocol; and GA-G: methodological and scientific analysis, drafting and correction of the protocol and linguistic aspects.
Funding This study was supported by a grant from the French Eastern Interregional Group of Clinical Research and Innovation (GIRCI Est) (Appel à Projet PARAmédical, APPARA 2022) and by CHRU de Nancy.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
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.