Objective To identify and analyse the interventions delivered opportunistically in secondary or tertiary medical settings, focused on improving routine vaccination uptake in children and young people.
Design Scoping review.
Search strategy We searched CINAHL, Web of Science, Medline, Embase and Cochrane Database of Systematic Reviews for studies in English published between 1989 and 2021 detailing interventions delivered in secondary or tertiary care that aimed to improve childhood vaccination coverage. Title, abstract and full-text screening were performed by two independent reviewers.
Results After deduplication, the search returned 3456 titles. Following screening and discussion between reviewers, 53 studies were included in the review. Most papers were single-centre studies from high-income countries and varied considerably in terms of their study design, population, target vaccination, clinical setting and intervention delivered. To present and analyse the study findings, and to depict the complexity of vaccination interventions in hospital settings, findings were presented and described as a sequential pathway to opportunistic vaccination in secondary and tertiary care comprising the following stages: (1) identify patients eligible for vaccination; (2) take consent and offer immunisations; (3) order/prescribe vaccine; (4) dispense vaccine; (5) administer vaccine; (6) communicate with primary care; and (7) ongoing benefits of vaccination.
Conclusions Most published studies report improved vaccination coverage associated with opportunistic vaccination interventions in secondary and tertiary care. Children attending hospital appear to have lower baseline vaccination coverage and are likely to benefit from vaccination interventions in these settings. Checking immunisation status is challenging, however, and electronic immunisation registers are required to enable this to be done quickly and accurately in hospital settings. Further research is required in this area, particularly multicentre studies and cost-effectiveness analysis of interventions.
- PUBLIC HEALTH
- Paediatric infectious disease & immunisation
- Health policy
Data availability statement
Data are available upon reasonable request. Further information on the studies included in this review can be obtained by contacting the corresponding author via email@example.com.
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
Our analysis and data synthesis have provided the first comprehensive overview of opportunistic interventions to improve uptake of routine vaccinations in secondary and tertiary medical settings.
We searched a large range of databases over an extensive time period and included studies from all around the world.
All data screening and extraction were performed by two independent reviewers.
We did not search the grey literature and may have inadvertently excluded interventions that are used in practice, or that failed to show benefit.
Only studies published in English were included.
Vaccination has made an enormous contribution to global health. Every year, immunisations save millions of lives and are one of the most successful and cost-effective public health interventions.1 Despite this, the UK, the USA and other countries with successful immunisation programmes experience outbreaks of vaccine-preventable diseases because of suboptimal vaccine coverage.2 Health inequalities exist in vaccination, with certain population groups more likely to experience poor coverage.3 The reasons for these inequalities are complex and influenced by a range of factors including:3
Vaccine hesitancy, due to:
Concerns about vaccine safety and efficacy.4
Misunderstanding around disease severity due to low incidence.5
Parental/carer resentment of perceived pressure to risk their child’s safety for population benefit.6
Mistrust of healthcare professionals (HCPs), governments and vaccine research.7 8
Reliance on unofficial information sources.7 8
Religious vaccination opposition (eg, Orthodox Jewish populations).9
Non-religious ‘anti-vaxx’ sentiment.10 11
Limited access to vaccines, due to:
Despite the success of COVID-19 vaccination programmes, evidence suggests that disruption caused by the pandemic has led to a global reduction in routine vaccination.16 For example, coverage of the first dose of human papillomavirus virus vaccine in UK females aged 12–13 years fell to 59.2% in 2019/2020, versus 88.0% in 2018/19% and 86.9% in 2017/18.17 UK childhood vaccinations are normally delivered in primary care settings; however, COVID-19 vaccination has demonstrated the suitability of alternative settings. Children and young people (CYP) can spend significant waiting time in secondary or tertiary care settings, which could be used to provide public health interventions. Indeed, the National Institute for Health and Care Excellence (NICE) recommends that the immunisation status of children be checked at every opportunity, including visits to the emergency department (ED), outpatient clinics and inpatient admissions, with vaccination either offered on the premises or referral to an appropriate vaccination service.18 NICE has also highlighted groups at risk of underimmunisation, including those with chronic illness or frequent hospitalisations, with secondary/tertiary care representing a key opportunity to vaccinate such children alongside the primary reason for their attendance.18
Maintaining vaccination uptake at levels required to prevent community disease spread may necessitate innovative approaches to vaccine delivery. This scoping review seeks to explore interventions delivered in secondary or tertiary medical care settings to improve routine vaccination uptake in CYP.
The scoping review question was:
What are the interventions delivered in secondary or tertiary medical care settings focused on improving routine vaccination uptake in children and young people?
We aimed to identify and analyse interventions to obtain a broad understanding of how they are delivered in hospital settings and their impact on routine vaccination uptake.
Throughout the review, the terms ‘vaccination’ and ‘immunisation’ are used interchangeably. Secondary care generally refers to treatment provided in hospitals, while tertiary care is for patients needing complex hospital treatment.21
The review considered studies that described interventions delivered in secondary or tertiary care to improve routine vaccination uptake among CYP published between 1 January 1989 and 11 October 2021. All countries were included. Interventions were considered opportunistic if they were not the primary reason for attending the healthcare setting.
As detailed in the protocol, we excluded studies not published in English.19
Search strategy and study selection
On 12 February 2020, we searched CINAHL, Web of Science, Medline and Cochrane Database of Systematic Reviews for articles published between 1 January 1989 and 12 February 2020, using search terms outlined in the protocol.19 The search was repeated and extended to include EMBASE on 11 October 2021. Duplicates were removed electronically, after which titles and abstracts were screened by two researchers independently before full paper retrieval. At each stage, disagreements were discussed, and consensus reached. Full papers were assessed against the inclusion criteria prior to data extraction and further discussion determined the final study sample. Conference abstracts were excluded due to insufficient information on the included interventions.
A data extraction form was developed using JBI guidelines to collect the information necessary for data synthesis (see online supplemental appendix 1). Two reviewers independently performed data extraction for all studies, with all authors involved at this stage.
Following data extraction, studies were tabulated by setting and publication date with intervention information presented alongside outcome data. Summary data were also extracted and tabulated based on key characteristics of the studies and interventions. Due to the varied nature of studies and interventions, no meta-analysis was performed.
Deviations from the protocol
Although the protocol stated that we would include children aged under 16 years, we also included studies with an older upper age range (up to 21 years) due to inability to extract data for younger children from these studies.
Patients and public involvement
No patients or public were involved.
In total, 12 421 titles were returned from the search strategy, after which 8985 duplicates were removed, leaving 3436 for title screening. After this, 342 records remained for abstract screening. Next, 148 full papers were retrieved and underwent full text review. Finally, data were extracted from 53 texts (figure 1). All stages were carried out by two independent researchers.
The included studies were extremely variable in terms of their population, target vaccination, clinical setting and intervention. Table 1 summarises the general characteristics of the included studies and associated interventions and online supplemental appendix 2 lists all included studies grouped by clinical setting and in chronological order.
General characteristics of studies
The studies were from 14 countries, predominantly the USA, Australia and other high-income countries. The most common settings were inpatient wards, followed by antenatal/neonatal settings, EDs and outpatient clinics. A range of age groups were examined in individual studies, with the most frequent being children of all ages, followed by younger age groups and four studies also including family members. Several vaccinations were studied, most commonly all due/overdue immunisations and influenza. Various study designs were used, encompassing quality improvement (QI) projects, clinical audits/service evaluations, cross-sectional studies, intervention studies, randomised controlled trials, cohort studies and pilot studies.
Characteristics of the interventions
Interventions varied substantially according to their content and delivery. Most involved predischarge vaccination and a third involved extra resources. Other common features were patient/family education, staff training/education, a multidisciplinary approach and the use of automatic vaccine ordering. The most common approach to checking immunisation status was parental/carer recall.
Note on settings
There were some considerations specific to setting, particularly neonatal settings. Here, several studies explored hepatitis B and BCG vaccination administered post birth. Although opportunistic in that it took place in hospital without appointment, this was often the recommended care setting for the vaccination. For example, national policy in the UK is for babies born to mothers with hepatitis B to receive vaccination within 24 hours of birth, usually in hospital.22
The pathway to successful opportunistic vaccination in secondary and tertiary care
The heterogeneity of the included studies illustrates that opportunistic vaccination represents a complex pathway and involves several steps to be successful, all with potential for patient drop-out. We have attempted to summarise this pathway below and provide a narrative summary of the approaches and interventions used at each stage:
Identify patients eligible for vaccination.
Take consent and offer vaccination.
Communicate with primary care.
Ongoing benefits of vaccination.
It should be noted, however, that not all interventions will encompass all steps; for example, educational interventions delivered in hospital, but where vaccination occurs in the community.
Identify patients eligible for vaccination
Baseline vaccination coverage
Several studies had assessed baseline vaccination coverage to determine the pool of eligible patients.23–46 For all due/overdue vaccinations, baseline coverage ranged from 44%24 40 to 89%,35 with little difference by setting and lower coverage in older studies. For influenza, baseline coverage was lower, ranging from 25%47 to 50.5%.31
Determining immunisation status
For vaccination to be successful, eligible patients must be accurately identified. This requires individual data, such as age, presence of underlying disease, immunisation status and clinical condition.
Checking immunisation status (henceforth referred to as ‘screening’) was most straightforward in neonatal studies where all infants were generally eligible. However, the complexity increased with age and cumulative number of required vaccinations. The target vaccination and setting were also important. As a single yearly vaccination, screening influenza vaccination status was more straightforward. In outpatient studies, patients had an ongoing relationship with the teams, reducing the complexity of screening, while inpatient stays afforded greater time to screen. Contrastingly, in ED there was limited time and rapid patient turnover.
In terms of personnel, screening was most successful in studies with extra staff and/or funding, including dedicated research staff.24 32–34 39 42 48–53 Elsewhere, there was no clear consensus regarding who was best placed for this task, although two studies had successfully used pharmacy staff.54 55
A range of methods were used to screen immunisation status.
Used in 30.2% of studies, this was the most common approach⇓.23 29–32 36–41 44 48 51 56 57 Although straightforward, it was inaccurate for studies of all due/overdue immunisations and was more appropriate for influenza. Szilagyi et al found that 20% of children reported as underimmunised in ED were actually up to date, while a quarter of those reported as up to date were underimmunised.39 When compared with immunisation registers, Ressler et al and Riley et al found that immunisation status based on recall was incorrect for 14.5% and 32.1% of patients, respectively.23 28
Electronic clinical alert system
These were used by 20.8% of all studies and involved influenza and hepatitis B vaccination.33 34 43 52 57–63 Systems were designed to generate automatic vaccination alerts, based on age and clinical risk factors. Alerts were often delivered alongside other digital initiatives, such as automatic ordering, or within wider QI initiatives. However, Pollack et al found automated screening to be a predictor of inpatient influenza vaccination uptake.46
Handheld immunisation documentation
This was used in 22.6% of studies, usually alongside other methods.24 26 29 30 36 37 39 40 44 48 56 The approach was unreliable, with Cunningham et al and Lindegren et al finding that 56% and 24%–26% of patients respectively had no documentation with them in ED.36 40
Phone calls to primary care
Two studies had screened immunisation status by telephoning primary care.24 30 This was inefficient, with Bell et al reporting an average of 1.5 calls to obtain a vaccination record and 4–5 hours spent daily calling primary care.24
Checking against a national or regional immunisation registry
This was the gold standard and most accurate approach. Two UK studies had combined checking handheld documentation with telephoning the local health authority to check registry data.23 25 Several Australian studies had used the Australian Immunisation Register, a national register that records all vaccines administered and which staff can access remotely.28 32 33 35 41 A New Zealand study had used a similar approach.42
Confirming clinical condition is compatible with vaccination
At this stage of the pathway, the patient’s clinical condition and any clinical contraindications must also be considered. Studies reported varying proportions of children too ill to be vaccinated, ranging from 0% to 20.5% and with no obvious relationship to setting.23 31 34 36 38 42 46 51 53 64 Leading reasons to defer vaccination were fever, diarrhoea, upcoming/recent surgery, vaccine allergies or oncology patients undergoing treatments.
Take consent and offer immunisations
Although clinical contraindications were important, vaccines not being offered and parent/carer refusal were greater contributors to non-uptake. Non-offer ranged from 11% to 77%, with the upper and lower range both in studies examining all due/overdue vaccines.25 37 42 44 No studies had evaluated why vaccines were not offered.
Many studies had explored parent/carer refusal of vaccination.25 31 37 38 42 46 48 51 56 60 65 This varied according to target vaccination and was low for neonatal vaccines, with Bakshi and Sharief reporting that 1% of parents refused neonatal BCG vaccination and Connors et al reporting that parental refusal was rarely or never a reason for not vaccinating against hepatitis B at birth.56 65 In contrast, refusal was higher for other vaccines in high-income countries. Here, for measles, mumps and rubella (MMR) or all due/overdue vaccines, parental refusal ranged from 9.4% to 37.5% where vaccination status was known, with Cunningham et al also reporting 87.5% refusal where status was unknown.25 37 38 40 42 For influenza, refusal ranged from 25.6% to 72% and was greater when offered in the ED.31 46 51 60 Across all vaccinations, four studies had evaluated underlying reasons, with common responses encompassing preference for vaccination in primary care, belief that the child was too unwell, concerns about the safety and efficacy of vaccination and belief that it was not needed for healthy children.31 40 46 47
Several studies of influenza and neonatal hepatitis B vaccination used automatic ordering/built-in order sets.43 46 54 57 58 61 65 66 Massey et al found that admission orders were associated with increased neonatal hepatitis B vaccination and Connors et al found that uptake of hepatitis B vaccination in a hospital where it was on a standing order was 93%–96% versus 71%–77% where it was not.66
Pharmacy involvement was frequently identified as essential to ensuring that vaccines were consistently available and dispensed quickly, with pharmacy staff involved in the leadership and delivery of several interventions.41 54 57 61 67 Gattis et al described a pharmacy-led intervention for influenza vaccination of solid organ transplant recipients whereby pharmacists were responsible for screening patients, assessing appropriateness, recommending vaccination to providers, educating patients/family and verifying and dispensing vaccines.55 Vaccination uptake rates increased from 36% pre intervention to 72% post intervention (p<0.001), with influenza diagnoses also falling.55
Next, vaccinations must be administered, with the potential for further drop-out. This was evidenced by Orenstein et al and Rao et al.54 who had evaluated how vaccine orders translated into administration, with only 40.3% and 61.2% of those with orders receiving vaccination, respectively.54 58
For each study, online supplemental appendix 2 summarises baseline coverage and subsequent outcomes, including administration and uptake of vaccination. Although uptake varied by study, virtually all demonstrated an improvement in coverage post intervention. It is difficult to compare administration rates due to variable study conditions and outcome measures; however, table 2 summarises ranges by setting and vaccination.
Looking first at influenza, higher uptake was generally seen in inpatients and outpatients than EDs, and in studies of children with underlying medical conditions.43 52 53 55 61 The highest uptake (90.32%) was reported by Lo and Sobota in an outpatient study of children with sickle cell disease.63 Similarly, Pappano et al and Rao et al found that underlying medical conditions were associated with increased vaccination.51 54
For all due/overdue vaccinations, there was higher uptake in studies with dedicated immunisation staff.33 35 36 39 40 48 49 Outside of these, intervention uptake was higher in older studies, with studies published pre-2000 reporting uptake of 65%–82.4% and those post-2000 reporting uptake of 3.4%–64%.
Uptake of neonatal vaccines was generally high. However, it was often unclear to what extent this was a consequence of the intervention, with little difference in outcomes pre intervention/post intervention or when compared with control. For example, Bolam et al reported 94% uptake of infant immunisations in the control group versus 96% in the intervention group.68 In studies of preterm and high-risk infants, however, interventions to increase uptake of routine vaccinations at chronological age through parental and staff education showed a marked improvement.64 69 70
Communicate with primary care
After vaccine administration, primary or community care providers must be informed. This was a further benefit of a remotely accessible vaccination registry, as used in Australia and New Zealand, with primary care updated of any vaccinations administered via this route.42
Communication with primary or community care was also important to arrange vaccination of children not vaccinated in hospital. In some studies, patients were referred to primary care if they were not vaccinated in hospital. In others, such as Muehleisen et al, there was no in-hospital offer of vaccination, with the intervention consisting of education and a prompt to arrange vaccination and primary care informed as such.26 Here, 27% of patients in the intervention group had received vaccination 1 month post discharge, compared with 8% of the control (p<0.001).26
Ongoing benefits of vaccination
Although not strictly part of the vaccination pathway, some studies had evaluated whether interventions had lasting impacts on coverage and vaccination behaviour.
Four studies had explored whether gains in coverage were sustained beyond the intervention’s initial timeframe (table 3). In the two examining all due/overdue vaccinations, initially increased uptake associated with the intervention was not sustained.26 39 Similarly, Kushner et al found that coverage of hepatitis B vaccination fell with time.59 However, in these studies it was unclear whether, in the absence of the intervention, coverage would have been even lower. In their study of all infant immunisations, Lemaitre et al found that uptake was consistently higher in the intervention group at all timepoints.71
As far as we are aware, this is the first attempt to review the literature relating to opportunistic vaccination across secondary and tertiary care settings and we have provided a comprehensive overview of interventions used to improve vaccination in these settings. Despite established childhood immunisation programmes internationally, there were relatively few published papers available. Similarly, although NICE recommends opportunistic vaccination in the UK, there were only four papers published between 1991 and 2007.15 23 25 56 Our review has demonstrated that opportunistic vaccination in hospital settings is complex, requiring several steps to be successfully navigated for interventions to be effective.
Vaccination coverage among CYP attending secondary and tertiary care appears to be below that of the general paediatric population.72–75 This was evaluated by some of the included studies, with, for example, Shingler et al reporting coverage of 70.6% in their study population versus a regional average of 85% and Tarca et al reporting coverage of 75% in their first study cohort versus a state and national average above 91%.35 42 This is important in the context of suboptimal uptake of many UK vaccinations, with only 85.3% having received the preschool booster and 86.6% the second MMR dose by age 5 in 2020/2021.76 With ongoing outbreaks of vaccine-preventable diseases, such as measles and pertussis, opportunistic vaccination in hospital-based settings may represent one route through which to vaccinate an underimmunised patient subgroup.77 78
An important finding was that, although the effect sizes were variable, virtually all interventions led to an improvement in coverage post intervention. This suggests that interventions were able to reach and vaccinate patients not vaccinated via traditional methods. Previous literature has shown that a key barrier to childhood vaccination is access, including time constraints, distance, location, long waiting times, childcare challenges for siblings and impermanent residence for groups such as homeless or looked-after children.79 Clearly, opportunistic vaccination overcomes these barriers and provides an opportunity to inform parent/carer knowledge about vaccination. Both Gilbert and Wrigley, and Conway reported that a leading reason for underimmunisation in the community were minor illnesses at the intended time of vaccination, as identified previously in the vaccination literature.25 29 79 Thus, hospital settings may present a useful opportunity to discuss true medical contraindications to vaccination and to vaccinate children in a setting where they can be monitored and their safety assured.
This review found consistent evidence that the effectiveness of opportunistic vaccination depends on the ability to quickly and accurately assess vaccination status, particularly for all due/overdue vaccines. National UK policy is for patients to be offered vaccines if their current vaccination status is unknown.80 However, Cunningham et al found that parents were reluctant to do so, with uptake of catch-up vaccinations in ED only 15% among patients with uncertain status compared with 71% with documented underimmunisation (p<0.0001).40 The review demonstrated that parental recall and handheld records were unfeasible screening options due to unreliability and unavailability, while confirming with primary care was time-consuming. Consequently, a remotely accessible electronic system is required to achieve this successfully, as demonstrated by studies using the Australian Immunisation Register.28 32 33 35 41 In the absence of this, inpatient admissions may be appropriate for catch-up of routine immunisations due to the prolonged time in hospital. Influenza vaccination may be possible in more time-pressured ED and outpatient settings due to the reduced screening required alongside the opportunity to use digital initiatives that reduce the burden on staff, such as electronic alerts and automatic vaccine ordering. This is especially relevant given that influenza vaccine uptake in the UK is lower than other childhood vaccines, with 56.7% uptake among 2 and 3 year olds in 2020/2021.81 In the UK, the National Health Service is transitioning to a digital handheld child health record (the ‘eRedbook’) from 2023, which may improve the long-term feasibility of opportunistic catch-up vaccination, although alternative short-term and medium-term interventions are likely to be required.82 83
Several studies described interventions that used additional staff and/or funding, which were generally more successful than those that did not. Even with digital interventions, delivering vaccination alongside routine care may be challenging without additional resources. In the study by Burgess et al, ED staff were reluctant to take on responsibility for vaccination and felt that they lacked sufficient time.38 Likewise, Cunningham et al described how, in the absence of the dedicated immunisation nurse, combining tasks with the existing duties of ED staff made immunisation a low priority, while Buenger and Webber reported that ED staff prioritised other tasks over influenza vaccination.40 57 In the inpatient setting, Walton et al found that over half of staff expressed concerns or considered inpatient vaccination inappropriate.27 Therefore, it is important that new interventions are adequately resourced, with implementation facilitated by staff education and QI methodologies to ensure that they become embedded within care.24 34 42 54 Additional factors limiting intervention success were high levels of parental refusal and non-offer of vaccination by staff. There has been extensive research into refusal of community-based vaccination; however, future work should seek to understand the specific barriers underlying parental refusal and non-offer of opportunistic vaccination in hospital settings.
The included papers provided variable information about the interventions, often with limited detail rendering evaluation difficult. In addition, most interventions had used several components making it difficult to draw out the impact of individual aspects. Most were single centre studies that reported on local initiatives and it is challenging to determine their wider generalisability. Although studies demonstrated improved vaccination coverage, none had evaluated cost-effectiveness and few had evaluated the medium-term/long-term impact of interventions. Nevertheless, NICE suggests that any intervention that improves vaccination coverage is usually cost-effective, particularly if it benefits underserved groups.18 We did not search the grey literature and may have missed interventions used in practice via this route—this also increases the risk of publication bias. Additionally, we only included studies published in English, potentially biasing findings towards those from English-speaking countries. As with all scoping reviews, we did not formally evaluate evidence quality and, due to the studies’ varied nature, only limited synthesis of results was possible.
This scoping review has explored and summarised the published literature relating to interventions delivered in secondary and tertiary settings focused on improving routine vaccination uptake in CYP, with most studies demonstrating improved vaccination coverage post-intervention. Furthermore, children attending hospital appear to have lower baseline coverage than the general paediatric population and are likely to benefit from interventions in these settings. For interventions to be successful, however, there is a need for electronic immunisation registers to enable vaccination status to be quickly and accurately checked, with the UK’s transition to the eRedbook a potential long-term route to facilitate this. Although existing research suggests that opportunistic vaccination interventions in hospital settings may be beneficial, further research is needed in this area, particularly multicentre studies and cost-effectiveness analysis.
Data availability statement
Data are available upon reasonable request. Further information on the studies included in this review can be obtained by contacting the corresponding author via firstname.lastname@example.org.
Patient consent for publication
As this review used published data only, no ethical approvals were required.
Contributors SB was involved in the design of the study, title, abstract and full-text screening, and data extraction. In addition, SB performed the analysis and data synthesis, wrote the first draft of the manuscript and undertook subsequent revisions to the manuscript. RE contributed to the coconception of the study, design of the study, the literature search and project management. RE was also being involved in title, abstract and full-text screening, data extraction, and the write-up of the manuscript at all stages. RI contributed to the coconception of the study, design of the study, title, abstract and full-text screening, and data extraction, along with supporting oversight of the project. RE and RI contributed equally to this paper and wish to be listed as joint last authors. KN, NG, LT, SS and MO contributed to data extraction. All authors were involved in critically revising the manuscript and approved the final version for submission. RE, RI and SB act as guarantors for this study.
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 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.