UNIT 8 ASSIGNMENT

Training paediatric healthcare staff in recognising, understanding and managing conflict with patients and families: findings from a survey on immediate and 6-month impact Liz Forbat,1 Jean Simons,2 Charlotte Sayer,3 Megan Davies,3 Sarah Barclay4

1Australian Catholic University and Calvary Health Care, Canberra, Australian Capital Territory, Australia 2Lullaby Trust, London, UK 3Evelina London Children’s Hospital, St Thomas’ Hospital, London, UK 4Medical Mediation Foundation, London, UK

Correspondence to Sarah Barclay, Medical Mediation Foundation, 36 Westbere Road, London NW23SR, UK; sarah.barclay@ medicalmediation.org.uk

Received 22 February 2016 Revised 23 March 2016 Accepted 25 March 2016 Published Online First 20 April 2016

To cite: Forbat L, Simons J, Sayer C, et al. Arch Dis Child 2017;102:250–254.

ABSTRACT Background Conflict is a recognised component of healthcare. Disagreements about treatment protocols, treatment aims and poor communication are recognised warning signs. Conflict management strategies can be used to prevent escalation, but are not a routine component of clinical training. Objective To report the findings from a novel training intervention, aimed at enabling paediatric staff to identify and understand the warning signs of conflict, and to implement conflict resolution strategies. Design and setting Self-report measures were taken at baseline, immediately after the training and at 6 months. Questionnaires recorded quantitative and qualitative feedback on the experience of training, and the ability to recognise and de-escalate conflict. The training was provided in a tertiary teaching paediatric hospital in England over 18 months, commencing in June 2013. Intervention A 4-h training course on identifying, understanding and managing conflict was provided to staff. Results Baseline data were collected from all 711 staff trained, and 6-month follow-up data were collected for 313 of those staff (44%). The training was successful in equipping staff to recognise and de-escalate conflict. Six months after the training, 57% of respondents had experienced conflict, of whom 91% reported that the training had enabled them to de-escalate the conflict. Learning was retained at 6 months with staff more able than at baseline recognising conflict triggers (Fischer’s exact test, p=0.001) and managing conflict situations (Pearson’s χ2 test, p=0.001). Conclusions This training has the potential to reduce substantially the human and economic costs of conflicts for healthcare providers, healthcare staff, patients and relatives.

INTRODUCTION Conflict is a recognised component of healthcare provision. Direct and indirect costs associated with conflict include litigation, reduced productivity, staff turnover and team morale.1 For patients, con- flict results in compromised decision-making2 and undermining trust in clinicians.3 4

Conflict consumes considerable amounts of staff time, particularly nurses and doctors.5

Communication difficulties are identified as a sig- nificant contributor to conflict,5 6 as are cross- cultural difficulties7 and religious beliefs.8

Underpinning each of these causes can be different

understandings of the clinical situation,9 different interpretations of futility10 and likely prognosis.11

In recent work documenting the incidence and severity of conflict in paediatric settings, the three most frequently cited causes of conflict between staff and patients/family members were: communi- cation breakdown, disagreements about treatment and unrealistic expectations.5

Conflict models offer ways of conceptualising strategies which may facilitate resolution. The Thomas–Killman’ two-dimensional model proposes that there is a need to balance assertiveness and cooperation,12 which includes further facets of col- laboration, competition, accommodation, avoid- ance and compromise. Although developed in the context of business and management-related con- flict, this model holds value in articulating core fea- tures of conflict management. The development of empathy, enabling the other party to maintain self- respect and self-esteem have been proposed as core elements of managing conflict.13 Growing recogni- tion of how conflicts develop and worsen facilitates awareness of when to intervene to minimise further escalation.3

While mediation may be a solution14 15 changing practice, focusing on staff understanding and

What is already known on this topic

▸ Conflict between staff and patients/families in paediatric hospitals can be a frequent and severe phenomenon.

▸ Direct and indirect costs associated with conflict include litigation, lower morale and reduced trust between staff and patients/ families.

▸ Empathy, communication and collaboration are recognised features in managing conflict.

What this study adds

▸ A 4 h tailored training programme increases staff ability to recognise conflict triggers and de-escalate conflicts.

▸ Staff reported that 6 months after the training, the focus on empathy and communication skills had led to changes in their practice.

250 Forbat L, et al. Arch Dis Child 2017;102:250–254. doi:10.1136/archdischild-2016-310737

Original article

team-management of conflict, may be more fruitful for man- aging emerging conflicts and early intervention. Training com- prises a core mechanism for changing how clinicians respond to the potential for conflict, particularly when the information can be used soon after the training.16 Yet paediatric trainees do not receive adequate conflict management training.17 Kaufman18

outlines a curriculum for teaching medical staff about identify- ing and responding to conflict, which takes account of time con- straints, the need for behavioural change, contextual power structures, assumed skills and the legal parameters of managing conflict. The paper concludes by stating a need for educational programmes to be tailored to meet these features.

This paper describes an innovative training course designed for staff in a paediatric hospital to recognise, manage effectively and de-escalate conflicts.

METHODS The training content was developed by SB and JS, based on understandings of conflict causes, impacts3 and severity5 in paediatric settings. The training mirrored Gerardi’s1 work on assessing the conflict, identifying some of the symptoms, under- lying causes and unhelpful assumptions which may exacerbate or cause conflict, and Back’s2 description of useful communica- tion tools such as active listening, empathising and self- disclosure. The training: (1) Provided information on what triggers conflict between

parents and health professionals and how to spot the warning signs,

(2) Included simulation exercises designed to encourage staff to empathise with patients and families by ‘stepping into their shoes’,

(3) Taught skills to help staff de-escalate conflicts with families.

The 4-h training sessions were run in multidisciplinary groups of up to 15 people, over the course of 18 months.

Training sessions began with an opportunity for participants to discuss in pairs a conflict they had experienced with a parent or patient, focusing on the impact of the conflict and on the thoughts and emotions they experienced at the time. The train- ing sessions also involved simulation exercises, asking partici- pants to play the part of a parent or health professional, or begin a conversation with a parent who is exhibiting the warning signs of potential or escalating conflict such as distress or anger.

Participants and measures Participant eligibility was determined by individuals being employed by the hospital (a tertiary paediatric hospital in England), at any grade or in any role. Nursing staff and non- consultant doctors were rostered to attend the training by their managers who encouraged participation of their teams and ensured that they were allowed time off from their clinical duties to attend. Training dates were also circulated via the Trust email so that any member of staff could apply for a training place. All staff who expressed an interest in the training were accommodated to attend.

A bespoke questionnaire was designed to determine the immediate and long-term impact of the training. The question- naire was administered at three time points: immediately before the training, immediately after and 6 months later. The first two were administered by paper copy. The third was sent via an elec- tronic survey to staff email addresses. Staff who were still working in the hospital were followed up in person by CS or MD. All questionnaires sought information on whether staff were able to recognise the triggers of conflict between families

and health professionals, and whether they had the strategies to manage conflict. Qualitative prompts asked participants to reflect on their main learning (‘tell us one thing you learnt from this training which you have found helpful in communicating with patients and their families’), and to record ‘any other com- ments you would like to make about the training and/or its impact on your practice’. Demographic information regarding staff role was also collected.

The 6-month follow-up survey collected data on: (A) whether the training had equipped staff to more readily recognise and de-escalate conflicts with patients and families, (B) describe one thing they had learnt from the training which they had found helpful in communicating with patients and their families, (C) whether they had experienced a conflict with a family since doing the training and if so, whether the training had helped them to (1) recognise the triggers and warning signs (2) to de-escalate or resolve the conflict.

Responses were recorded on either a 5-point Likert scale, or as a simple yes/no binary. Analysis was primarily conducted using descriptive statistics, to enable reporting of percentages, mean, mode and median scores. Respondent identifiers were not used, prohibiting treating responses as paired data. Pearson’s χ2 and Fischer’s exact tests were used to examine a priori hypotheses (significance set at p=0.05) regarding the impact of the training on ability to recognise signs and triggers from pre training to 6-month follow-up and on differences between nursing and medical professionals. Data were organised as fre- quency counts and percentages of people who answer in each

Table 1 Study participants

Staff group Baseline/post 6 months

Administrator 11 5 Chaplaincy 1 4* Clinical nurse specialist 35 18 Consultant 28 15 Manager 5 6 Matron 5 3 Non-consultant doctor 87 15 Nursing assistant 9 10 Other 61 34 Paediatric nurse practitioner 9 4 Staff nurse 368 156 Therapist 27 16 Ward sister 65 27 Total 711 313 % of total trained 100 44

*In some staff categories there are greater numbers at 6 months than baseline/post assessment. We believe that some respondents coded themselves as ‘other’ at first assessment and then identified differently at follow-up.

Table 2 Quality and relevance of training

Quality of training n (%) Relevance n (%)

Excellent 506 (71.2) Very relevant 609 (85.7) Good 181 (25.5) Relevant 92 (12.9) Satisfactory 11 (1.5) Not relevant 2 (0.3)

Poor 1 (0.1) Missing data 8 (1.1) Missing data 12 (1.7) Total 711 (100) 711 (100)

Forbat L, et al. Arch Dis Child 2017;102:250–254. doi:10.1136/archdischild-2016-310737 251

Original article

Likert category at each time point, to report observed and expected frequencies.

Qualitative data collected from free-text prompts were ana- lysed drawing on thematic analysis, adopting a five-stage process of familiarisation, identifying a thematic framework, indexing the data, synthesising across respondents and data interpretation to form key themes.19 Analysis was informed by a position of theoretical freedom, rather than a priori hypotheses regarding the likely content or themes arising from the data.20 Analysis was conducted by an experienced qualitative researcher, with discussion of emergent themes with the wider team.

The study was conducted in one tertiary paediatric teaching hospital in England. Data collection commenced in June 2013 and ceased on 30 May 2015, with the training provided from June 2013 until November 2014. This study was deemed by the hospital’s Research and Development team to be service evalu- ation and consequently was not reviewed by a health service research ethics committee.

RESULTS Seven hundred and eleven staff were trained and completed baseline data, 313 of whom completed questionnaires at 6-month follow-up. Table 1 provides details of respondents’ staff role and the number of completed surveys at each time point.

Staff rated the quality of the training very highly, with 98.5% rating it excellent or good, and 99.8% rating it very relevant or relevant as indicated in table 2.

Participants were asked about their ability to recognise trig- gers for conflict and use of skills to manage conflicts. Table 3 summarises the binary yes/no responses and illustrates an

improvement from baseline to immediate-post training assess- ment. Table 4 illustrates the observed and expected frequencies across the Likert scale for recognising triggers. Fisher’s exact test indicated a significant difference between the scores 4 and 5 in the pretraining responses compared with the 6-month follow-up data (43.7% vs 57.8 and 6.7% vs 29.1%, p=0.001). Figure 1 illustrates changes in staff ability to recognise triggers from baseline to 6-month follow-up

Table 5 demonstrates the observed and expected frequencies across the Likert scale for pre and 6-month follow-up data. Pearson’s χ2 test indicated a significant difference between the scores 4 and 5 in the pretraining responses and the 6-month follow-up responses (20.1% vs 58.5 and 3.1% vs 17.9% respectively, p=0.001). Figure 2 illustrates changes in staff ability to deal with conflict from baseline to 6-month follow-up.

At 6-month follow-up, participants were asked if the training had equipped them to recognise and de-escalate conflicts with patients/families. The majority (n=283, 90%) reported that the training had had this impact.

Six months after the training 178 staff respondents (57%) had experienced conflict. Of those 178, 169 (95%) said that the training had enabled them to recognise the triggers for the con- flict. One hundred and sixty-two (91%) reported that they had also been able to de-escalate the conflict as a consequence of the training.

Data from baseline and 6-month follow-up were analysed to determine if there were differences between nurses’ and doctors’ responses to the training, in reporting scores of 4 or 5 (able or very able) to recognise and deal with conflict. Neither analysis approached significance at baseline or follow-up (recognise con- flict, p=0.459; deal with conflict, p=0.725). Consequently, the training appeared to have comparable impact across staff groups.

Analysis of the qualitative data identified five core themes, and a further six minor themes. The five core themes were: communi- cation and listening, recognising warning signs/triggers, improve- ments in practice, empathy and perspective taking. Participants identified that being aware of early warning signs and triggers was key learning from the training, impacting practice:

The training has been a key factor in the fact that I have not experienced any conflicts in the last few months. Early recogni- tion of triggers has helped me avoid conflict developing. (Clinical nurse specialist)

The training was so useful! Our department faces conflict daily. The training came into use three times the day after the course. All three were potentially explosive situations which I felt very able to manage. I think little updates/refreshers to the training would be most valuable. (Paediatric dental specialist)

Table 3 Learning about identifying and managing conflict

I can:

Baseline (n=711)

Post (n=711)

6 months (n=313)

n (%)

Recognise the triggers of conflict between families and health professionals.

Yes: 349 (49) No: 354 (50) Missing data: 8 (1) Mean: 3.5

Yes: 682 (96) No: 20 (3) Missing data: 9 (1) Mean: 4.4

Yes: 272 (87) No: 41 (13) Mean: 4

Use appropriate skills and strategies for recognising and dealing with conflict at different levels of severity.

Yes: 163 (23) No: 540 (76) Missing data: 8 (1) Mean: 3

Yes: 640 (90) No: 62 (9) Missing data: 9 (1) Mean: 4.2

Yes: 239 (76) No: 74 (24) Mean: 4

Table 4 Ability to recognise triggers

Ability to recognise triggers scores 1: not very able–5: very able

Time period 1 2 3 4 5 Total

Pretraining 6-month follow-up

Count 3 29 316 307 47 702 Expected count 2.1 21.4 245.5 337.5 95.4 702 % within time period 0.4% 4.1% 45.0% 43.7% 6.7% 100.0% Count 0 2 39 181 91 313 Expected count .9 9.6 109.5 150.5 42.6 313 % within time period 0.0% 0.6% 12.5% 57.8% 29.1% 100.0%

Total Count 3 31 355 488 138 1015 Expected count 3.0 31.0 355.0 488.0 138.0 1015 % within time period 0.3% 3.1% 35.0% 48.1% 13.6% 100.0%

252 Forbat L, et al. Arch Dis Child 2017;102:250–254. doi:10.1136/archdischild-2016-310737

Original article

Many staff reported specific strategies from the training which they were using routinely, reinforcing the positive impact of the practical nature of the training. One such strategy derived from the training, but not explicitly suggested to participants, was to manage the environment in which difficult conversations took place, for example, moving a parent from the ward to a room to enable a more private conversation. In response to the prompt ‘One thing I learnt was…’ staff offered the following responses:

The ability to remove a parent from a tense environment to a side room where she/he may be able to express himself/herself in confidence and without interruption and to be listened to actively. (Staff nurse)

Effective listening, and not hesitating to apologise and not give false hope. (Staff nurse)

Not confronting them, but allowing the patient/relative to vent their frustrations and focus your efforts on understanding the cause of their frustrations rather than denying or opposing their views. (Non-consultant doctor)

Learning how to develop an empathic approach by ‘stepping into the shoes’ of patients and families was reported by many respondents as having a profound impact on their approach to engaging with them:

[The training] made me try to put myself in the shoes of patients and their relatives, and to think about things from their perspec- tive much more. (Consultant)

To see it from the families’ perspective more. Even if I may not fully agree with the argument/issue I now empathise more with the stressful situations the families are in. (Staff nurse)

DISCUSSION This tailored training, delivered to staff in a paediatric hospital, resulted in a significant improvement in the ability to identify and manage conflict with patients and relatives. Unresolved conflict over goals of care that escalate may require external interventions such as independent mediation or court inter- vention.21 22 This training therefore has the potential to reduce the need for such costly and stressful involvement of third parties.

Evidence-based methods of addressing conflict are required, since conflict regarding treatment and goals of care is a marker for increased risk of complicated bereavement for families23 and is an independent predictor of burn-out in staff.24 Previous research has demonstrated the impact of conflict management training on reducing employee stress25 and consequently points to the potential for positively impacting morale.26 The impact on the quality of care has yet to be established. Indeed, conflict can be construed positively as a way of energising and prompt- ing initiation of new conversations to manage hostility.27 The findings support other calls for training on conflict management to be built into healthcare infrastructure.28

Although simulation training has been criticised,29 the blended approach to this training, including role play, appears to have had a substantial positive impact on attendees’ self-reported ability to recognise and then manage conflict

Table 5 Ability to deal with conflict

Ability to deal with conflict 1: not very able–5: very able

Time period 1 2 3 4 5 Total

Pretraining 6-month follow-up

Count 19 156 365 140 22 702 Expected count 14.5 109.3 300.9 223.4 53.9 702 % within time period 2.7% 22.2% 52.0% 19.9% 3.1% 100% Count 2 2 70 183 56 313 Expected count 6.5 48.7 134.1 99.6 24.1 313 % within time period 0.6% 0.6% 22.4% 58.5% 17.9% 100%

Total Count 21 158 435 323 78 1015 Expected count 21.0 158.0 435.0 323.0 78.0 1015 % within time period 2.1% 15.6% 42.9% 31.8% 7.7% 100%

Figure 1 Ability to recognise triggers. Figure 2 Ability to deal with conflict.

Forbat L, et al. Arch Dis Child 2017;102:250–254. doi:10.1136/archdischild-2016-310737 253

Original article

situations with patients and families. The use of self-report measures is a recognised methodological weakness.30 31 The lack of control or comparison group, for example using other communication skills development approaches32 or compas- sion,33 which are known to reduce patient distress,34 compro- mises claims about this intervention being superior to other communication interventions. Since participant identifiers were not used in any of the survey cycles, paired analysis was not possible. This limits the ability to track individual transform- ation over the training and follow-up timeline. The lack of paired data also precluded fine-grained analysis of which staff groups’ responses reflected greatest levels of reported change; this is particularly salient since conflict is not experienced uni- formly across staff groups.5

The training was offered only on a one-off basis, and further evaluation should be conducted on the additive value of a refresher course. Further evaluation could include additional measures to record impact on staff performance, such as impact on number of conflicts experienced, changes in family/patient satisfaction with care (to allow for comparison with staff reports of being able to de-escalate 91% of conflicts), alongside mea- sures to report any impact the training had on intrastaff conflict.

The study focused on conflict between staff and patient and families, and consequently did not examine intrastaff conflict. Some of the training may have had a positive impact on this, but it was not measured. Loss to follow-up at 6 months may be partially explained by some staff (notably doctors in training) no longer working at the hospital and therefore being less engaged in the ongoing evaluation. Respondents returning questionnaires at 6 months may be a skewed sample of those most satisfied or highly impacted by the training, despite identical reminders from two members of the team.

CONCLUSION With an established need for interventions which help manage conflict5 this training provides an evidence-based approach to training healthcare staff. The training has the potential to reduce the human and economic costs of conflict, by furnishing staff with the appropriate skills and knowledge to identify and then de-escalate potential and actual conflicts.

Twitter Follow Liz Forbat at @lizforbat

Contributors SB, JS and LF designed the work. CS and MD acquired the data. LF, CS, MD and SB interpreted the data. LF and SB drafted the work and revised it critically for intellectual content. LF, JS, CS, MD and SB approved the final version of the manuscript. LF, JS MD, CS and SB agree to be accountable for all aspects of the work ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding The study was funded by the Guy’s and St Thomas’ Charity (Grant: EFT120609).

Competing interests SB received a grant from the Guy’s and St Thomas’ Charity, during the conduct of the study; and she is the director of the Medical Mediation Foundation—an organisation which provides conflict management training and mediation in situations where there is disagreement/conflict between patients and healthcare professionals. However, the manuscript focuses on conflict incidence not mediation as a solution.

Ethics approval Not required.

Provenance and peer review Not commissioned; internally peer reviewed.

Data sharing statement Any requests for raw data should be directed to the corresponding author.

REFERENCES 1 Gerardi D. Using mediation techniques to manage conflict and create healthy work

environments. AACN Clin Issues 2004;15:182–95. 2 Back AL, Arnold RM. Dealing with conflict in caring for the seriously ill: “it was just

out of the question”. Jama 2005;293:1374–81. 3 Forbat L, Teuten B, Barclay S. Conflict escalation in paediatric services: findings

from a qualitative study. Arch Dis Child 2015;100:769–73. 4 Chan TC, Bakewell F, Orlich D, et al. Conflict prevention, conflict mitigation, and

manifestations of conflict during emergency department consultations. Acad Emerg Med 2014;21:308–13.

5 Forbat L, Sayer C, McNamee P, et al. Conflict in a paediatric hospital: a prospective mixed-method study. Arch Dis Child 2016;101:23–7.

6 Azoulay E, Timsit JF, Sprung CL, et al. Prevalence and factors of intensive care unit conflicts: the conflicus study. Am J Respir Crit Care Med 2009;180:853–60.

7 Twiss SB. On cross-cultural conflict and pediatric intervention. J Relig Ethics 2006;34:163–75.

8 Brierley J, Linthicum J, Petros A. Should religious beliefs be allowed to stonewall a secular approach to withdrawing and withholding treatment in children? J Med Ethics 2013;39:573–7.

9 Ashwal S, Perkin R, Orr R. When too much is not enough. Pediatr Ann 1992;21:311–7.

10 Bluebond-Langner M, Belasco JB, Goldman A, et al. Understanding parents’ approaches to care and treatment of children with cancer when standard therapy has failed. J Clin Oncol 2007;25:2414–19.

11 Paris JJ, Schreiber MD, Moreland MP. Parental refusal of medical treatment for a newborn. Theor Med Bioeth 2007;28:427–41.

12 Thomas K. Conflict and conflict management. In: Dunnette M, ed. The handbook of Industrial and organizational psychology. Chicago: Rand McNally, 1976:889–935.

13 Halpern J. Empathy and patient-physician conflicts. J Gen Intern Med 2007;22:696–700.

14 Ward Platt M, Ward Platt A. Conflicts of care. Arch Dis Child 2005;90:331. 15 Meller S, Barclay S. Mediation: an approach to intractable disputes between parents

and paediatricians. Arch Dis Child 2011;96:619–21. 16 Murdoch-Eaton DG. Managing difficult issues: does training have any impact? Arch

Dis Child 2013;98:1–2. 17 Ahmed A, Moore H, Purva M, et al. What do senior paediatric trainees want

from simulation in Yorkshire: training needs assessment survey. Arch Dis Child 2013;98:A11–A13.

18 Kaufman J. Conflict management education in medicine: considerations for curriculum designers. Online J Workforce Educ Dev 2011;5:1–17.

19 Braun V, Clarke V. Successful qualitative research: a practical guide for beginners. London: Sage, 2013.

20 Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Research in Psychology 2006;3:77–101.

21 O’Brien A, Sokol DK. Lessons from the Ashya King case. BMJ 2014;349:g5563. 22 Richards T. When doctors and patients disagree. BMJ 2014;349:g5567. 23 Neimeyer RA, Burke LA. Complicated grief and the end-of-life: risk factors and

treatment considerations. In: Werth JL, ed. Counseling clients near the end of life. New York: Springer, 2012:205–225.

24 Embriaco N, Papazian L, Kentish-Barnes N, et al. Burnout syndrome among critical care healthcare workers. Curr Opin Crit Care 2007;13:482–8.

25 Haraway D, Haraway III W. Analysis of the Effect of Conflict-Management and Resolution Training on Employee Stress at a Healthcare Organization. Hosp Top 2005;83:11–17.

26 Saltman DC, O’Dea NA, Kidd MR. Conflict management: a primer for doctors in training. Postgrad Med J 2006;82:9–12.

27 Sui H, Spence Laschinger HK, Finegan J. Nursing professional practice environments: Setting the stage for constructive conflict resolution and work effectiveness. Journal of Nursing Administration 2008;38.

28 Young A, Turner J. Developing Inter-professional Training for Conflict Resolution—A Scoping Audit and Training Pilot. Mental Health Review Journal 2009;14:4–11.

29 Norman G, Dore K, Grierson L. The minimal relationship between simulation fidelity and transfer of learning. Med Educ 2012;46:636–47.

30 Agrawal S, Norman GR, Eva KW. Influences on medical students’ self-regulated learning after test completion. Med Educ 2012;46:326–35.

31 Eva KW, Regehr G. Self-assessment in the health professions: a reformulation and research agenda. Acad Med 2005;80:S46–54.

32 Egan G. The Skilled Helper—a problem management approach to helping. California: Brooks Cole, 1998.

33 Fogarty LA, Curbow BA, Wingard JR, et al. Can 40 seconds of compassion reduce patient anxiety? J Clin Oncol 1999;17:371–9.

34 Maguire P, Pitceathly C. Key communication skills and how to acquire them. BMJ 2002;325:697–700.

254 Forbat L, et al. Arch Dis Child 2017;102:250–254. doi:10.1136/archdischild-2016-310737

Original article