Article Text
PDF
Abstract
Objective To describe the pharmacists’ workflow, including tasks and time spent, to better understand their work capacity.
Design Cross-sectional, observational, time and motion study.
Setting Community pharmacies in Western Australia and New South Wales, Australia.
Participants Currently registered and practising pharmacists were approached using snowball sampling and selected using purposive techniques to obtain balance representation of metropolitan and rural pharmacies, as well as high and low script volumes where possible.
Results Twenty-four pharmacists across 15 pharmacies participated during the 135 sessions totalling over 274 hours of observation. Dispensing (30%), indirect patient services (17%), counselling (15%) and professional management activities (15%) were the top four duties pharmacists performed, while only 2% of time was spent on professional services such as pain clinics and influenza vaccinations. Tasks were frequently interrupted and often performed simultaneously. Breaks and consumer-contact times were limited. More time was spent on professional service activities in non-metropolitan pharmacies, in pharmacies with greater daily prescription volumes and those with one or more support pharmacists.
Conclusions This is the first study to quantify the pharmacists’ tasks in Australian community pharmacies. Much time is being spent on dispensing, supply and management activities with little time for providing additional professional services. An extra supporting pharmacist is likely necessary to increase professional services. These findings could support future research around barriers and enablers of conducive workflows and of extended professional services.
- primary care
- organisation of health services
- health services administration & management
- quality in health care
- change management
Data availability statement
Data are available upon reasonable request. Deidentified participant data are available from the corresponding author on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- primary care
- organisation of health services
- health services administration & management
- quality in health care
- change management
Strengths and limitations of this study
This is the first large-scale, time and motion study quantifying the pharmacist’s work patterns in Australian community pharmacies.
The study captures over 274 hours of direct observation.
Recruitment of participants from both metropolitan and rural pharmacies and across two states increases data generalisability.
Direct observational approach may change the behaviour of pharmacists being observed.
Background/introduction
Community pharmacists predominately dispense medicines and advise on medication matters.1 However, funding pressures, growing and ageing population and health workforce shortages have led to calls for community pharmacists to provide a wider range of consumer-centred, clinically orientated professional services (PS) such as medication reviews and chronic disease management.2 3 The resultant services aim to complement general practitioner (GP) and allied health service offerings and help reduce public demand for services in overloaded emergency departments and medical clinics.4 There are many definitions for these health services within the pharmacy sector and they are differentiated from dispensing and related professional activities by terms including ‘cognitive’, ‘extended’, ‘enhanced’, ‘newer’, ‘patient-centred’, ‘professional’ and ‘clinical’ pharmacy services.3–5 In this paper, such non-dispensing-related activities and non-over-the-counter medicine sales services, and those identified by the Australian Seventh Community Pharmacy Agreement (7CPA) as patient-centred services will be referred to as ‘professional services’.5
In Australian community pharmacies, PS such as vaccinations, Diabetes MedsCheck, Quality Use of Medicines Maximised for Aboriginal & Torres Strait Islander people and health check programmes are already offered.3 5 Globally, PS have shown to identify and resolve medication-related problems and improve health outcomes, yet a number of barriers have affected their adoption and implementation in Australia.6 7 Recent community pharmacy–government agreements have looked to address concerns around remuneration models being more product-orientated (fees for dispensing each item) rather than for patient-centred services that possibly need extra time and staffing.8–10 Every 5 years, the Commonwealth government negotiates community pharmacy funding with the Pharmacy Guild of Australia and the Pharmaceutical Society of Australia.5 The latest 7CPA has specified remuneration of $1.2 billion to support professional programmes and services (in addition to $14 billion for dispensing and over $1 billion for wholesalers) from July 2020.5 While this is likely to increase provision of PS, a number of additional barriers have been recognised, suggesting that the uptake of pharmacy services may be more complex than simple financial remuneration. For example, appropriately trained staff and consultation rooms are required to incorporate new services into the pharmacy workflow.11 Fragmented GP–pharmacist relationships, differing consumer perceptions on whether pharmacists should/could offer other services, consumer demand, pharmacist knowledge, confidence and attitudes, increasing workloads and challenging work areas (busy, limited space, noisy) are other potential determinants.2 9 12–14
One important challenge consistently voiced by pharmacists15 is the perceived lack of time and inadequate support staff numbers that hinder the more time-consuming pharmacist-led PS.16 17 Current Australian literature18 19 is limited to just providing an overview of the daily duties of community pharmacists, and there is limited quantitative evidence that supports these concerns. No study has measured the time spent on specific tasks, nor the overall capacity within the pharmacists’ workday to find time for additional services.10 Documenting time pressures across a range of community pharmacist’s workdays is critical to better understand whether pharmacists have the capacity to extend roles within their current practice and business model.
Compared with previous studies using work sampling techniques20 and qualitative interviews and focus groups,21 quantitative, time and motion methodology has been suggested as the more appropriate approach to obtain data on work patterns. It has been used to describe pharmacist work in hospital settings,22 23 and non-Australian hospital and community pharmacies.12 21 23 24 However, to our knowledge, time and motion methodology in Australian community pharmacy has only been used in one pilot study by Cavaye et al.10
This study is part of a larger project designed to extend Cavaye and colleagues’ work10 by increasing the number of participating pharmacies and covering both urban and rural areas. The objective of this paper is to describe the work of community pharmacists including key tasks performed by pharmacists, and how their time is spent during the workday, with the goal of understanding the potential capacity for pharmacists to include PS in their daily practice.
Study design
The detailed study protocol has been published previously.2 The Space for Pharmacy-based Interventions and Consultation Time study was a quantitative, cross-sectional, community pharmacy-based observational study that used the Work Observation Method By Activity Timing (WOMBAT) tool25 26 (figure 1) to collect time and motion data on pharmacist workload allocation across time, during routine workdays.
Screenshot of Work Observation Method By Activity Timing (WOMBAT) data collection software tool. P Comm, professional communication; P Edu, professional education; P Mgmt, professional management; P Service, professional service.
Patient and public involvement
Community pharmacists were the target population for the study. As a result, patients or the public were not involved in this research. However, outcomes may benefit the general public by gaining an understanding of the pharmacists’ duties and their role within the primary healthcare system. The results of this study will be disseminated through various means including published papers and presentations.
Recruitment
Community pharmacists from across regional and metropolitan areas of Western Australia (WA) and New South Wales (NSW) were invited to take part in the study over a 15-month period between October 2018 and December 2019. Pharmacies were classified geographically as per the Modified Monash (MM) model on a scale of MM1 (major city) to MM7 (very remote).27 Pharmacies were purposively approached to obtain representation from a balance of metropolitan and rural pharmacies as well as high and low script volumes to maximise the generalisability of the results. A snowball technique28 was used to identify potential candidates from our own networks and knowledge of local areas and then asking them to suggest other community pharmacists. Pharmacy owners allowing observations in their pharmacy and each volunteering participant pharmacist provided written informed consent. Participating pharmacies nominated one or more consenting pharmacists to complete the planned three 6-hour days of observations. Each participant was an active, registered, practising pharmacist (ie, not just undertaking managerial duties, and not an intern/student pharmacist). Observed pharmacists were given a $50 voucher as reimbursement.
Data collection
Six observers, four from WA and two in NSW, were trained in the use of the WOMBAT software and task definitions (table 1). The multidimensional WOMBAT tool has been used to capture health professional communication and work patterns in previous studies and it allowed researchers the flexibility to create and tailor the necessary observational variables in this study (figure 1).23 26 Each observer started data collection when an inter-rater reliability score of above 85% had been obtained29 with another data collector who simultaneously, but independently, observed the same pharmacist.
Summary of tasks and definitions2 10
Observations were purposefully scheduled across a variety of weekday shifts, between the hours of 08:00 and 18:00 to maximise data variability and capture the full pharmacy workday. An observer equipped with the WOMBAT software26 loaded on an Android tablet shadowed the pharmacist participant in 2-hour blocks, limited to three sessions per day, each separated by a short break to minimise participant and observer fatigue. Observation timings were thereafter changed to acquire work patterns during all business hours. Time data were collected across four categories, including: what task, with whom, where and mode (summarised in table 1 and figure 2), that were defined using descriptions formulated in an earlier study.2 10 The tool also captured whether the pharmacist was interrupted or multitasking.26 30
Diagram of pharmacy layout used during data collection to indicate locations of task completion detailed in table 1.
The observer was stationed in close proximity to the pharmacist, yet far enough away to avoid interruptions to pharmacist’s duties and maintain privacy during conversations. Consumers were informed of the study through posters positioned around the pharmacy and if requested, in person by the pharmacy team. Observers avoided interactions with consumers unless requested by the pharmacy, and no consumer or identifying data were collected. The observer did not record any conversation or information inadvertently witnessed or not related to the study.
Data analysis
WOMBAT data were downloaded and analysed in MS Excel 2016 and STATA MP V.16.0 (Statistics/Data Analysis; Texas, USA, www.stata.com) as per WOMBAT analysis guide V.2.2.25 Work patterns were identified through calculations of the time spent on work tasks and multitasking, and the rate of interruptions. Time spent was reported as a proportion of the total observation time, or total task-specific time (ie, total time spent of the task noted) or of the total variable-specific time (eg, ‘whom’ variables such as alone and with consumer; and ‘how/mode’ variables such as face to face and on a computer). Other descriptive statistics including the frequency, median and IQR of different tasks and common task combinations were also calculated. Additionally, the relationships between the PS task occurrence and pharmacy and pharmacist type, number and type of supporting staff and prescription items to sales ratio were cross-tabulated to describe how these impact PS provision.
Results
Demographic details
Fifteen community pharmacies accepted the invitation to participate from a total of 18 approached. All 15 were included and provided data for the final analysis (online supplemental appendix 1). Challenges, in particular the time taken for pharmacies to respond to invitations and/or sign paperwork, and consequent impact on timelines and budget, prevented recruitment of the originally planned 20 pharmacies.
Supplemental material
Nine pharmacies were recruited in WA: five metropolitan (classified as MM1: major cities accounting for 70% of Australia’s population) and four non-metropolitan (two classified as MM2: regional with over 50 000 residents, and two MM3: rural with between 15 000 and 50 000 residents).27 The remaining six were in metropolitan areas of NSW (MM1: major cities).27 Metropolitan sites were a mixture of five single/co-owned independents, two franchisees and four large commercial chain pharmacies. Non-metropolitan sites consisted of one pharmacy chain and three franchisees. Prescription volumes, as estimated by pharmacy owners, ranged from 300 to 2680 prescriptions per week, with nine pharmacies (all from metropolitan areas) estimating volumes below the Australian average of 1035 prescription items per week.31 Staffing per pharmacy ranged from having 0 to 4 additional pharmacists (ie, not including the observed pharmacist, median=0, IQR: 0, 1), 0 to 16 other pharmacy staff (trained in the supply of medicines under pharmacist supervision, median=2, IQR: 1, 7) and up to 11 non-pharmacy (ie, sales, median=0, IQR: 0, 1) staff. Generally, staffing levels were in line with or sometimes better than those recommended by the Australian dispensing practice guidelines32 of having at least one trained dispensary assistant for dispensing workloads between 150 and 200 items per day, and additional pharmacists or dispensary assistants if dispensing levels exceed 200 items per day.
A total of 12 male and 12 female pharmacists participated: 1 locum, 10 pharmacist owners/managers and 13 pharmacists (online supplemental appendix 1). One metropolitan-based pharmacist was specifically employed as a PS manager in charge of conducting clinical services. Nearly 70% (n=16) were below the age of 40 years and a similar proportion (n=16) had been qualified for up to 15 years. The majority (88%) were Australian citizens (18 in the metropolitan area and three in non-metropolitan pharmacies), and 75% (n=18) also attained their pharmacy qualification in Australia. All but three pharmacists worked on a full-time basis (n=21, 88%). More than one pharmacist was required in eight of the participating sites (four WA, four NSW) to fully complete the 18 hours of observations.
Work patterns
Data collection took place over a total of 135 sessions and 274 hours:40 min:40 s of observations (‘total observation time’), with a mean time of 18 hours:18 min:43 s per pharmacy. A total of 7028 tasks summing to 293 hours:52 min:5 s (‘total all-tasks time’) were recorded (online supplemental appendix 1). The ‘total all-tasks time’ is greater than the actual ‘total observation time’ because it includes overlapped, multitasking times (when two or more tasks were conducted concurrently). Consequently, proportions of total observation time (predominately used in this study) do not add to 100%.
Tasks
Dispensing, indirect patient services (non-face-to-face activities for consumers, including ordering or organising of medication supply and delivery), counselling and professional management activities occupied over 80% of the total observed time (33.0%, 16.6%, 15.0% and 14.5%, respectively, figure 3). PS were only observed on 31 occasions, summing to 1.8% of the total observation time. Table 2 includes the time spent on each task and their quartile ranges. The median time per task for any task performed was 1 min and 22 s (IQR: 39 s, 2 min:47 s). Lengthier times per task were logged during compounding (median=4 min:24 s per task, IQR: 1 min:59 s, 9 min:51 s) and PS tasks (4 min:7 s, IQR: 2 min:32 s, 10 min:22 s), and the least amount of time on communication activities with pharmacy staff and/or other health professionals/prescribers (‘professional communication’ 41 s, IQR: 19 s, 1 min:38 s).
Distribution of task time, multitasking and interruptions
Tasks observed, where performed, with whom and mode of task completion as a percentage of total observation time (n=274 hours:40 min:40 s).
Location and interactions
Pharmacists spent the greatest proportion of their time at locations where consumer interactions were limited. Over 53% of time was recorded in the dispensary and 9% in the back office or outside of the pharmacy compared with almost 38% of time at the pharmacy counter, on the shop floor or in the consult room (figure 3). The dispensary was the primary location for dispensing, compounding, indirect patient services, waiting, and professional communication, education and management activities. General counselling and sales tasks were chiefly performed at the pharmacy medicines counter. Online supplemental appendix 2 details the most common combinations by task type of location, persons interacting with and mode.
Supplemental material
The multiselectable ‘with whom’ category logged pharmacists working alone for nearly 48% of their time and interacting with pharmacy and non-pharmacy workers/professionals for 13% of time each. Pharmacist–consumer interactions were witnessed for 35% of the total observation time: 10% in the dispensary or back office combined and the remaining 25% at the pharmacy counter, shop floor and consult room combined. Consumer communications were predominately around general counselling (42.0% of total consumer-contact time), dispensing (32.6% of total consumer-contact time) and sales tasks.
Multitasking
Pharmacists performed two or more tasks simultaneously during 7.5% (over 20 hours) of their observed time (table 2). They were more likely to multitask during indirect patient services (9.6% of total task time) and dispensing activities (9.4%), followed by professional education, sales and counselling tasks (6.8%, 6.2% and 6.1% of respective total task times). Indirect patient services were mostly multitasked with professional communication tasks, and dispensing events with counselling.
Interruptions
As detailed in table 2, pharmacists were interrupted on 845 occasions at a rate of three interruptions per hour or had one interruption every 19 min on average. Professional education activities were interrupted the most (one per 9 min), in comparison to dispensing and indirect patient service activities that were both interrupted once per 14 min and professional management tasks that were interrupted once per 17 min. Interruptions typically occurred to undertake professional communication duties, and less frequently to undertake counselling and dispensing jobs.
Professional services
PS observations such as blood pressure measurement, blood glucose checks, pain clinics and influenza vaccinations were minimal. Only 31 observations and 4 hours:54 min (1.8% of the total observation time) were recorded (table 2). However, the median time per observation of 4 min:7 s (IQR: 2 min:32 s, 10 min:22 s) was greater than the other tasks. Nine out of 15 pharmacies performed at least one PS task, with half (50.5%, 2 hours:28 min:26 s) of all observations being registered at just three non-metropolitan sites and the remaining spread across six metropolitan pharmacies (four in NSW and two in WA). These tasks were predominately conducted face to face in available consult rooms (for 73.8% of total PS task time).
Pharmacies with greater PS times (median=2 hours:22 min, IQR: 1 hour:13 min, 3 hours:40 min) tended to have larger prescription volumes (median=164 prescriptions per day, IQR: 114, 292) and more medicine dispensing transactions than the total number of sales transactions (86% of total task time, 4 hours:12 min). Male pharmacists registered greater times spent on PS tasks compared to females (70% of total task time, 3 hours:25 min), as did pharmacies with at least one other supporting pharmacist (77% of total task time, 3 hours:48 min). However, it was noted that these were all based on a small number of PS observations. Pharmacists performing these tasks were less likely to be interrupted (one per 147 min) or have activities multitasked.
Discussion
This is the first large-scale direct observational time and motion study documenting work patterns in Australian community pharmacies. Pharmacists were busy, undertaking 25 tasks per hour on average with a median of 1 min:22 s (IQR: 39 s, 2 min:47 s) per task, interspaced with infrequent or minimal breaks (1 per hour, median=41 s (IQR: 19 s, 1 min:38 s)) and repeated interruptions (one every 19 min). Pharmacists spent 80% of their time on dispensing, counselling duties, management tasks such as organising of staffing rosters and indirect patient services (activities related to the supply of prescription medicines: clarifying orders and sourcing medicines). While such product-focused tasks are necessary and essential to the pharmacy business, this leaves little time for more consumer-focused, advanced PS. All participating pharmacies listed a selection of additional PS on offer; however, only a handful were observed to actually provide them during this study, accounting for less than 2% of the observed time.
This study had similar findings to previous studies conducted in the UK,33 Portugal,12 the USA24 and in an Australian feasibility study,10 suggesting some uniformity across countries. All found that high levels of product-focused workloads, being constrained to the dispensary, frequent multitasking (averaging two episodes every hour in this study) and short breaks (of around 4 min/hour in this study), may limit pharmacists’ choice of work.10 12 24 33 A literature review by Lea et al33 explored this concept further and reported that such work pressures and lack of opportunity to extend to more consumer-centred roles increased stress levels and reduced job satisfaction. Additionally, Tucker et al34 and Davies et al35 suggest that rest breaks of less than around 10 min/hour during prolonged continuous work may increase the chance of making errors. Therefore, in Australian pharmacies, a change in organisational processes aimed at having a more streamlined work pattern and frequent breaks may improve work fulfilment and preserve patient safety.33–35
Our work supports common perceptions of suboptimal work conditions and limited work capacity of Australian community pharmacists to expand roles.3 9 15 33 36 Barriers to adoption and implementation listed in previous studies of lack of time, generally high workloads and tasks dominated by dispensing and medicine supply are consistent with our data.9 33 Findings in this research of PS tasks (registering 1.8% of total observed time) being among the least documented align with Davies and colleagues’ pharmacists’ time utilisation study using work sampling methodology, showing that such tasks occupied just 3.2% of the pharmacists’ time.35 A study by Hattingh et al and a narrative review by Buss et al concluded that having access to private areas within pharmacies, and sufficient and skilled staff could help facilitate community pharmacy services.8 11 However, our study documented private consultation areas in almost 75% of the participating sites and pharmacy staffing numbers as recommended in professional guidelines,32 yet we observed relatively few PS tasks. A shift to exploring other important factors driving uptake of such services is required.
This study also found that the pharmacies with at least one additional support pharmacist, greater prescription volumes and non-metropolitan pharmacies provided more PS. However, as the proportions of pharmacies conducting PS were too small in this study, these findings cannot be generalised. Further time and motion research, possibly using a purposive sample of pharmacies who engage in regular PS, is recommended to provide additional clarification on specific influential factors. Time constraints are a commonly cited barrier to PS delivery.4 16 37 Other literature infers that non-metropolitan pharmacies provide more patient-orientated services due to having lighter workloads and, hence, more time; or such areas being underserved and, hence, placing greater emphasis on these pharmacists to extend their roles.37 38 Yet, a narrative review by Howarth et al challenges the validity of these results suggesting limitations in study design and inadequate control of potential confounding variables.39 Consequently, future qualitative research using pharmacist interviews and/or focus groups could supplement our quantitative findings and further elicit and explain work patterns in those providing PS, and between metropolitan and non-metropolitan community pharmacy practices.
While there is ample evidence of benefits from pharmacist-led PS, results presented here, though limited in sample size, suggest that practical implementation lags behind this evidence.3 5 9 12 35 Lack of adequate funding has been noted as a common barrier.40 The current Australian, volume-based, fee-for-service model only provides payments for each prescription dispensed or when a selected service is delivered.40 Time away for alternative roles could possibly be financially counterproductive. Jackson and Urick recommended a performance-based payment model that has been used by GPs, other allied health professionals and in other countries.40 In this model, payments could be incentivised according to patient health outcome targets such as medication compliance. Additionally, Chan et al detail the importance of including costs for training, participation in accreditation processes and documentation time in the remuneration package to ensure the viability of professional pharmacy services.41
However, finance may not be the only factor influencing the uptake of additional pharmacy services. As pharmacies differ in their organisation, geography and settings, the issue of identifying determinants is possibly more complex and multifactorial, and approaches require further and ongoing investigation.33 41 42 Solutions such as using other trained non-professional staff to undertake some medicine supply-related tasks; negating competing priorities by, for example, allowing more time for direct patient care rather than for documentation; addressing pharmacist training needs; and improving GP–pharmacist collaboration have been suggested.9 15 21 24 41 43 Gregório and colleagues also propose using internet-based pharmaceutical services to reduce unnecessary pharmacy visits and following examples of nursing practices that reorganised their work to accommodate similar services.12 This all warrants further exploration. It is also important to examine how the findings in this study are perceived by pharmacists, pharmacy staff, primary care teams and consumers to better understand the views of what the pharmacists’ role is, should be and how it could be better managed.33
Strengths and limitations
The strengths of this study included the length of observed time as well as the diverse nature of the pharmacies enrolled. There are, however, limitations to our data. First, there was the possibility of purposeful behaviour changes by participants (Hawthorne effect) knowing they were being observed.44 However, this was minimised with the observer sitting a distance away and as unobtrusively as possible, while maintaining long-sustained observational periods as recommended in other time and motion studies.12 35 45 Second, there was the possibility of interobserver variance in recording activities and observer bias. For example, three sites recorded only few or no multitasking and interruption observations relative to others that had around three times more. However, extensive training and regular conversations on data collection techniques between observers and study investigators, as well as formal inter-rater reliability measures to ensure consistency prior to commencing data collection were undertaken to mitigate variabilities. Finally, pharmacy recruitment challenges limited our ability to recruit the broadest range of pharmacy range of pharmacy types as planned and thus the generalisability of our findings to all pharmacy models is uncertain.
Conclusion
The evidence presented here is the first quantitative data set detailing work of community pharmacists in Australia. Pharmacists’ work choices are limited by time pressures, frequent interruptions and multitasking. Dispensing, general counselling and management activities seem to dominate daily duties while PS are among the least performed. If pharmacists are to expand their role in line with the 7CPA, a review of practice guidelines and organisational processes is required by policymakers and pharmacy owners. In light of these findings, more research into remuneration models, staffing structures, consumer preferences and primary healthcare professional perceptions of community pharmacist roles are the likely next steps.
Data availability statement
Data are available upon reasonable request. Deidentified participant data are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the University of New South Wales and Curtin University Human Research Ethics Committees (HREC number HRE2018-0635). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We wish to thank the contributions of all employed observers and participating pharmacists as well as the pharmacy owners who consented and allowed us to collect data in their pharmacies.
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
Twitter @Robinsonsuz
Contributors AK, RW, RN, SR, EL, T-LL, ID, CB and RJ contributed to the conception and design of the study. AK, RW, SR, RN and CB are involved in implementation of the study. AK coordinated the project and data collection, and drafted the manuscript. EL, ID, CB, T-LL and RJ critically reviewed the manuscript for important intellectual content. AK, RN, SR and RW revised and finalised the manuscript. All authors read and approved the final manuscript. All authors state that they had complete access to the study data that support the publication. AK is responsible for the overall content as the guarantor.
Funding This study is supported by a National Health and Medical Research Council (NHMRC) Project grant (ID1084657). AK is supported by the Australian Government Research Training Program Scholarship. T-LL is supported by the NHMRC Early Career Fellowship (APP1110230). RJ is supported by the National Heart Foundation Future Leader Fellowship (ID100484) and by a UNSW Scientia Fellowship. EL is supported by the Swedish Research Council for Health, Working Life, and Welfare (Forte), Junior Research Grant (2016-00148). RW is supported by an NHMRC Early Career Fellowship (APP1125044).
Competing interests None declared.
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.