Article Text

Download PDFPDF

Preventing or accelerating emergency care for children with complex healthcare needs
  1. David Sutton1,
  2. Paul Stanley1,
  3. Franz E Babl1,2,
  4. Fiona Phillips1
  1. 1
    Emergency Department, Royal Children’s Hospital, Melbourne, Victoria, Australia
  2. 2
    Murdoch Childrens Research Institute and University of Melbourne, Melbourne, Victoria, Australia
  1. Franz E Babl, Paediatric Emergency Physician, Clinical Associate Professor, University of Melbourne, Emergency Department, Royal Children’s Hospital, Parkville, Vic 3055, Australia; franz.babl{at}rch.org.au

Abstract

Objective: A subgroup of children with special health care needs (CSHCN) have chronic and complex medical conditions and frequently attend the emergency department (ED). Some of these ED visits could be prevented through appropriate clinician advice or, if an ED visit is unavoidable, the management time could be decreased. We set out to determine whether an ED-based advice and coordination programme was feasible and could prevent or accelerate ED care for these patients.

Methods, setting and patients: We identified CSHCN who frequently attended the ED at a large tertiary children’s hospital. These patients were enrolled in an ED-based coordination programme, the Accelerated Care through Emergency (ACE) programme providing 24-hour mobile-phone access to experienced ED nurses. We prospectively tracked usage patterns and determined the rate of ED visits after receiving phone advice and the waiting time for patients to be seen in ED. Parental satisfaction and cost of the programme were also assessed.

Results: After a pilot phase in 2002, enrollment in the programme increased from 125 in 2003 to 220 patients in 2006. Patients had a broad range of medical conditions. All had two or more and up to 22 medical services involved in their care. 80% of patients used a technical device or implant. Phone calls increased from an initial average of 31 per month in 2003 (0.24 calls per participant) to 66 per month in 2006 (0.3 calls per participant), 60% of which were after hours. The percentage of ED reviews per phone call dropped from an initial 74.2% (95% CI 55.2%–88.1%) in 2003 to 50.0% (95% CI 37.4%–62.5%) in 2006 (p = 0.02). However, decreases in ED visits and admissions as a percentage of enrolled patients and as a percentage of phone calls to ACE staff were not statistically significant. Mean waiting time for enrolled patients remained below 30 minutes. Parent satisfaction with the programme was rated 8.3 on a 0–10 scale (0 meaning poor, 10 meaning excellent). The approximate cost of the programme per child was AU$750 (£292) per year.

Conclusion: We have developed a coordinated approach towards the provision of healthcare for a group of families with diverse severe chronic medical conditions who frequently present to the ED. Through a comprehensive programme including the development of patient-care plans, care coordination and 24-hour mobile-phone access we were able to enhance families’ capacities to manage their children’s conditions in the community.

Statistics from Altmetric.com

Request Permissions

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.

Care for children has changed dramatically because of medical advances in neonatal intensive care, surgery, pharmacology and the use of sophisticated medical devices. These medical advances can, however, be associated with a significant effect on quality of life. For example, 29% of babies with extremely low birth weight who survive into adolescence are moderately or severely disabled.1 Although the definitions vary, children with special healthcare needs (CSHCN) make up a sizable proportion of children in any country. In the US, CSHCN make up 18% of children on the basis of a broad definition of children with chronic conditions that require health and related services of a type or an amount beyond what is required by children generally.2 3 In Australia in 2001, 44% of children aged 0 to 14 years were reported to have a long-term condition (defined as any condition lasting 6 months or more), with 25% having two, and 18% having three or more long-term conditions.4 In 2003, 8% of all children were reported to have a disability using the Australian Institute of Health and Welfare definition of a disability as something that restricts a child’s ability to perform tasks associated with daily living.4

The need to improve support for the families caring for these children has been recognized. In the US the concept of the medical home for CSHCN has been embraced by the American Academy of Pediatrics5 and the US Maternal and Child Health Bureau.6 This has led to a number of initiatives and models to improve community-based care for CSHCN.7 8 In the UK9 and in Australia,10 11 the need to address similar issues has been raised. Neville9 suggests that specialist children’s hospitals should be primarily geared to provide appropriate support for disabled children.

Children with chronic conditions frequently present to paediatric EDs and subsequently require admission to hospital at high rates (24% to 38).12 13 A subgroup of these patients repeatedly present to the ED with a range of medical problems.14 Although the acuity and complexity of their problems often forces these families to attend tertiary EDs, these patients are then often dealt with by staff unfamiliar with their extensive histories and their diverse and sometimes specialized needs. In addition, prolonged waiting times and exposure to pathogens from other sick children can be particularly problematic for these children. In light of this, it could be proposed that some of the ED presentations of CSHCN could be either managed at home or safely deferred to a planned or urgently arranged appointment with the primary service provider or an appropriate medical or allied health specialist.

On the basis of recurrent anecdotal parent, patient, ED and subspecialty staff feedback and a formal parent survey we decided to improve emergency care for patients with chronic and complex illnesses and/or disabilities. A literature review did not provide obvious solutions or programmes to either prevent or, if an ED visit is unavoidable, to accelerate emergency care for these patients. In this article we describe an ED-based nurse-led programme to achieve these goals and track the usage patterns of the enrolled families. We monitored the rate of phone contacts, ED visits, inpatient admissions and waiting times for patients enrolled in the programme and seen in the ED as the main outcome measures. In addition, parent satisfaction was measured. We also attempted to determine the cost of the programme and the number and types of patients who would potentially benefit from such a programme in a given population at a tertiary centre. Many of the findings may be applicable and valuable in similar settings.

METHODS

Programme development and structure

The Accelerated Care through Emergency (ACE) programme was developed at the Royal Children’s Hospital (RCH), a 260-bed tertiary children’s hospital in Melbourne, Australia. RCH has 34,000 inpatient admissions and an annual ED census of 56,000 patients. An advisory group was formed in 2001 that included parents, ED nursing and medical staff and subspecialty representatives as well as the RCH division of home and community care. We were funded by a Victorian Government Department of Health pilot grant, and we identified a group of experienced paediatric ED nurses (ACE nurse coordinators) who would provide 24-hour on-call mobile-phone coverage on a part-time basis. ACE nurses were all required to have extensive ED triage and resuscitation-room experience and be at least Paediatric Life Support certified. After an orientation programme for ACE staff, the programme was set up to create a patient database, develop clinical pathways and individual patient-care plans, in conjunction with subspecialty services involved. The database, which is available on computers in the ACE offices as well as on personal digital assistants (PDA) for the ACE nurse coordinators, includes patient demographics, names and contact details for all medical and non-medical specialists involved in their care, community supports, medial conditions and medications and specific care instructions and alerts.

ACE nurse coordinators assessed patient problems over the phone. Where appropriate they offered advice and support to enable families to manage their child at home or, if necessary, triage the patient to community or hospital services. If families were managed at home, ACE staff provided phone follow up and liaised with appropriate specialists by phone or email. If families were referred to the ED, ACE nurses would streamline the ED process by communicating directly with the senior ED physician/nurse and with the ED triage staff about the patient’s current problem and needs. In this way medical records and necessary hardware, such as replacement gastrostomy or nasogastric tubes, were pre-ordered; similarly, organizing a bed for hospital admission before arrival could be carried out if it was deemed necessary. The ED computer log system was configured to flag enrolled ACE patients with electronic alerts and highlight care needs. The ACE patient database was available on ED computers. ED policies were changed to encourage review of ACE patients by senior ED staff for rapid decision making. In general, all patients were contacted within 24 hours of the initial phone contact to ensure quality of care, ascertain feedback and update care plans. Medical supervision for the ACE nurse coordinators was through primary service or subspecialty consultants for individual patient-related issues and though ED consultant staff for urgent clinical issues.

Patient enrollment

Initially, we identified patients who attended the ED four or more times over a 12-month period and had chronic health conditions. After discussion with the subspecialty services involved, ACE nurse coordinators approached the parents to enroll them in the programme in 2002. After the initial pilot group, patients were referred to the ACE programme mostly through hospital subspecialty care teams, but also through allied health staff and medical social workers. Occasionally, parents self-referred to the ACE programme.

All referred patients were assessed for their suitability with the patient’s primary-care team and a review of the child’s medical record and presentation pattern to the ED. Parents/carers were then contacted for verbal consent to participate in the programme and were provided with written information about the programme and contact details. Programme staff attempted to involve the family’s community supports and general practitioner wherever possible.

Data collection and definitions

A data-collection sheet was designed for use in abstracting data during telephone calls. Data collected included date and time of call, reason for the call, advice given, disposition and follow-up arrangements. The ED log was reviewed daily for ED visits and hospital admissions by ACE programme patients. Staff interviews, electronic log data and medical record review were used to determine ED management including waiting times. Primary diagnosis was based on the consensus judgement of the authors as to the most important underlying diagnosis. Technical and implanted devices were defined as all such devices excluding hearing aids or glasses. After hours was defined as outside weekdays 0900–1700. ED waiting time was defined as time from presentation at triage to being seen by a medical staff member. Time to discharge or decision to admit was determined to be the time from triage to the time when a patient was discharged or when a medical staff member asked senior nursing staff to request a hospital bed. ACE programme coordinators classified calls as leading to “unavoidable ED presentations” if patients had to be seen in the ED setting for procedures related to device malfunction such as replacement of nasogastric or gastrostomy tubes. Phone calls to ACE nurse coordinators that resulted in patients avoiding an ED presentation, or from families who were seeking advice on their child’s management, were also recorded. The main measure of success of the ACE programme was a reduction in the number of ED visits per month per enrolled patient and per phone call.

Data from an anonymous parent satisfaction survey, undertaken in 2004, are also presented. The survey consisted of a mix of open and closed questions and was mailed to all enrolled patients at the time. Parent satisfaction was assessed on a 10-point scale from 0 (poor) to 10 (excellent).

Finally, we also present data from a scoping process undertaken in 2004 to determine the number of potential ACE patients not enrolled at the time. Frequent ED attendees were analysed over a 12-month period in terms of demographics, ED presentations and admissions.

The study was approved as an audit by the Human Research Ethics Committee.

Data analysis

Data were entered into the ACE electronic database. We analysed the data using descriptive statistics, confidence intervals and chi square testing using Stata software (Version 9.0, Stata Corporation, College Station, TX, USA). Values of p⩽0.05 were considered statistically significant.

RESULTS

Epidemiology of the ACE programme

After the pilot phase, the number of enrolled ACE patients increased steadily and plateaued at 220 patients in 2006. Figure 1 shows the number of enrolled patients, averaged in 6-month intervals, from January 2003 to June 2006 as a balance of families newly enrolled and patients either discharged from the programme (eg, to adult services or because of a change in the patient’s condition) or after the patient had died. In total, approximately 500 patients were enrolled in the programme at any one time. An analysis of the first 100 patients enrolled in 2002 indicates that by 2006 28% remained in the programme, 48% had been discharged from the programme, 2% had moved interstate, 6% had died and 14% were lost to follow up. In 2006 the average age of patients was 8.1 years (range 3 months to 18 years) (Table 1). Although there were many conditions and hospital services were involved, 44% had underlying neurological problems and 40% were cared for primarily by the child development and rehabilitation (CDR) service. All children had at least one other medical service or ancillary service involved in their care, 40% having more than five and in one case up to 22 additional services involved. Owing to the complexity of their medical conditions (often children had multiple implanted technical aids), CDR patients had the highest number of additional services involved. Eighty percent of enrolled children used at least one of a broad range of technical device or implant and 50% used at least two (Table 1).

Table 1 Patients Enrolled in the Accelerated Care through Emergency (ACE) programme
Figure 1 Development of the Accelerated Care through Emergency (ACE) programme. * Average per month, see text.

The demographics of patients in 2003, including the number of primary diagnoses and services involved, were broadly similar to the distribution in 2006 (data not shown).

Telephone calls to ACE coordinators, ED visits and admissions

Patient phone calls to ACE nurse coordinators increased from an average of 31 (0.24 calls per enrolled participant) to 66 (0.3 calls per enrolled participant) per month (Figure 1) from 2003 to 2006. Sixty percent of parent phone calls to ACE coordinators occurred after hours. Figure 1 also shows the average monthly number of ED presentations and hospital admissions in enrolled patients, averaged in 6-month intervals. The number of ED presentations as a percentage of enrolled patients remained unchanged with 23 of 125 per month (18.4%, 95% CI 12.0%–26.3%) in 2003 and 33 of 220 (15.0%, 95% CI 10.5%–20.4%) in 2006 (p = 0.41). The number of hospital admissions as a percentage of enrolled patients with 10 of 125 per month (8.0%, 95% CI 3.9%–14.2%) in 2003 and 16 of 220 per month (7.3%, 95% CI 4.2%–11.5%) in 2006 similarly remained unchanged (p = 0.67). The admission rates after ED presentation were 43.5% (95% CI 23.2%–65.5%) in 2003 and 48.5% (95% CI 30.7%–66.4%) in 2006 (p = 0.7).

The percentage of ED reviews per phone call dropped from an initial 74.2% (95% CI 55.2%–88.1%) in 2003 to 50.0% (95% CI 37.4%–62.5%) in 2006 (p = 0.02). The percentage of ACE calls leading to a hospital admission was 32.2% (95% CI 16.7%–51.4%) in 2003 and 24.2% (95% CI 14.5%–36.3%) in 2006 (p = 0.4).

ED visits deemed by ACE nurse coordinators as “unavoidable” are shown in Figure 2. This includes presentations for procedures such as replacement of nasogastric or gastrostomy tubes which owing to technical complexity or expertise could not be performed by parents or carers at home. The monthly average remained fairly steady at five to eight such presentations. Figure 2 also shows “avoided ED presentations”, where ACE nurse coordinators judged to have avoided an ED presentations through a telephone conversation. Avoided ED presentations increased over the programme period from eight per month to 30 per month.

Figure 2 Accelerated Care through Emergency (ACE) programme: Avoided and unavoidable emergency department presentations. ED, emergency department. *Unavoidable ED presentations: ED presentations for procedures such as replacement of nasogastric or gastrostomy tubes. Avoided ED presentations: phone call to ACE nurse coordinators judged to have avoided ED presentations.

ED waiting times

ED waiting times (averaged over 6-month intervals) for ACE-enrolled patients, whether admitted to hospital or not, from presentation at triage to review by an ED staff member, remained unchanged below 30 minutes ranging from 19 minutes to 25 minutes over the course of the programme (Figure 3). For ACE patients the time averaged over 6-month intervals from presentation at triage to discharge or decision to admit or discharge by a medical staff member ranged from 58 minutes to 147 minutes. Whereas the time to discharge or decision to admit was unchanged over the study period, for admitted patients the time from presentation at triage to leaving the ED for a hospital ward ranged from 134 minutes to 217 minutes with an increasing trend between 2003 and 2006. The latter time interval is influenced by the availability of inpatient hospital beds, a factor beyond the control of the ACE programme.

Figure 3 Waiting times in the emergency department for patients enrolled in the Accelerated Care through Emergency (ACE) programme. ED, emergency department. *Waiting time: time from presentation at triage to being seen by a medical staff member. Presentation to ward: time from presentation at triage to leaving ED to ward. Presentation to decision: time from presentation at triage to discharge or decision to admit by a medical staff member.

We elicited parent feedback in 2004 using a single mailing of an anonymous survey to all enrolled patients at the time (n = 160) with a return rate of 47%. Almost all families had at some time contacted the ACE nurses, some frequently, mainly for advice or because there was a major change in their child’s condition (Table 2).

Table 2 Accelerated Care through Emergency (ACE) programme parent satisfaction survey (n = 75)

Advice by ACE nurses was described as easy to obtain and thought to be helpful. Parents were in general very satisfied with the service (8.3 on a 0 to 10 scale). About half the parents (55%) thought that contact with the ACE nurses had avoided an ED presentation. When asked how ACE could be improved, 44 of 55 (80%) thought that there was nothing to improve. Three families suggested acceleration of admissions, one family suggested improved ACE phone access and another expansion of ACE to other hospitals.

Number of potential ACE patients and cost of the programme

In 2004 we undertook a scoping process to determine the broad demographics, ED presentations and admissions of all patients not in the ACE programme who had four or more unplanned ED visits during one year. On the basis of a review of electronic ED and hospital data we identified 193 such patients who had a total of 1124 ED visits resulting in 863 admissions (71%) over the preceding 12-month period. Approximately 50 patients from this group met the ACE criteria of long-term chronic health conditions and frequent ED presentations and did not have an existing detailed care plan. Twenty-eight of these patients, with similar characteristics to previously enrolled patients, were eventually enrolled in the programme. The remaining 143 patients belonged to three groups. (1) Predictable treatment needs in the ED that were or could be managed through existing treatment protocols. It included 94 oncology patients who had a comprehensive care plan with multifaceted support and communication links. It also included 18 patients with recurrent visits for chronic asthma who were part of a comprehensive asthma outreach programme. (2) Patients with short-term periods of frequent admissions due to recurrent illness associated with infancy, illnesses that took some time to diagnose or manage and palliative conditions. (3) Frequent attendances in the ED without associated hospital admissions, often associated with psychosocial issues.

On the basis of 220 enrolled patients, annual cost of the programme (salaries, telephone services and IT support) at a total of AU$166,000 (£64,740) translates into an average cost per patient of AU$750 (£292) per year.

DISCUSSION

We developed a new ED-based programme to identify and enroll patients with chronic and complex illnesses who frequently attend the ED. The original impetus to base this programme in the ED and provide staffing through ED nurses was that these patients presented to the triage desk with problems that could frequently be handled with advice from and coordination efforts by ED staff who often knew the patients and their care needs. In addition, administrative issues surrounding ED presentations and coordination with subspecialty services often impeded or delayed the provision of necessary clinical interventions. Many of these issues could be addressed or prepared for before the patient presented to the ED.

Through the development of patient-care plans, care coordination and 24-hour mobile-phone access to care coordinators who are experienced ED nurses, the programme enhances the families’ capacity to manage their children’s conditions in the community. The programme evolved into a virtual medical home for CSHCN fulfilling many of the goals set out by the American Academy of Pediatrics.5

The patients enrolled in the programme clearly represent the severe end of the spectrum of complex chronically ill children: all were cared for by more than one service, in more than 40% of cases more than five services were involved. Eighty percent of patients had at least one, and in 30% of cases at least two medical devices or implants. The admission rate of ACE patients after ED presentation (45%) was higher than rates reported in other studies of ED presentations of children with chronic conditions in the US (24%)13 or Belgium (38%).12 Although the effect of this programme is not easy to measure, we found that the programme proved popular with parents and referring subspecialty services as the increase in enrollment figures shows. The programme data presented also show that a number of calls by families could be addressed by providing phone advice and support, therefore avoiding ED presentation. The percentage of calls handled in this way had increased over time (p = 0.02), which is likely to be because of increasing familiarity of ACE staff with the patients and increasing parental trust in the quality of the advice given and a commensurate increase in their confidence to handle problems at home. Although parental feedback indicated that more than 50% of parents felt that ACE phone calls helped to avoid ED visits, the decreases in the numbers of ED visits per month per enrolled patient and per phone call were not statistically significant. A US programme to coordinate care for CSHCN at paediatric practices through nurse-practitioner-based case management and care plans did not lead to a decrease in ED visits.8 Data from adult EDs indicate that multidisciplinary case management does not reduce ED visits in frequent users of EDs.15 16 For many of these adult patients, the ED was thought to provide the most appropriate place to access acute medical care, offering a 24-hour service, without exclusion criteria, no monetary charge and a safe environment with high levels of medical and allied health resources.15 17

If an ED visit could not be prevented, the ACE programme was intended to speed ED assessment and discharge or admission. ACE patients were seen quickly and average ED waiting times remained under 30 minutes throughout the programme period. In addition to preparing the ED for the patient through advance contact with ED staff by the ACE nurse coordinators, other efforts to ensure speedy assessment by ED staff included automatic “flagging” of ACE patients though electronic alerts on presentation to triage, policy changes to provide for patient review by more senior ED staff and the availability of computer-based care plans. The development of parent-held patient record or electronic records for CSHCN presenting to EDs had been proposed in the past.13 18 The time to discharge or making of a decision regarding admission remained unchanged during the study period. It would be expected that ACE patients would have longer assessment times in the ED than other patients owing to their complex management needs and the need to coordinate with multiple services. The increasing ED dwell time for admitted patients from the time when a decision to admit had been made until transfer to an inpatient ward was disappointing; however, this probably reflects increasing demands on bed availability throughout the hospital system. During the study period the percentage of ED patients overall with a length of stay greater than 12 hours in the ED increased 2.5-fold from 0.6% to 1.5% over the study period and was to a large degree outside the control of ACE or ED staff.

Parental feedback was in general very positive, and ACE nurse coordinators were perceived as helpful by 96% of parents who responded. Two thirds of families used the ACE programme one to five times over a 12-month period although some parents used the programme much more frequently. The programme is also highly regarded by subspecialty medical teams, and it has been embraced by the ED staff. The ACE programme has also become a focal point for communication from community-based services involved in service provision or care for ACE patients such as the ambulance service.

Programme data indicate that the majority (60%) of the phone calls to ACE nurse coordinators occur after hours. Staffing and resource provision need to reflect this pattern. On the basis of the scoping process outlined above to determine the number of patients who fulfil ACE criteria (four or more ED visits per year in patients with chronic and complex illnesses), an ACE-type programme at an institution such as ours would require resources for about 250 patients. Considering the complexity and severity of this group of patients and an average cost of AU$340 (£133) for an ED visit without costs for tests (data from the finance department, Royal Children’s Hospital), the resources required for the programme of AU$750 (£292) per patient per year are modest. Although the calculations are based on subjective judgements by staff, “avoided” ED visits (30 per months × cost of ED visit) add up to savings of AU$122,000 compared with the annual cost of the programme of AU$166,000. Australia has universal, taxpayer-funded healthcare coverage. ACE programme funding was initially through a Victorian state health department government pilot grant. Ongoing funding is through Victorian government and internal hospital funding.

There are a number of areas that could be improved. ACE nurse coordinator supervision is currently ad hoc through ED consultants, or for specific patient-related questions through the primary hospital consultants in charge of the patient. Ideally the programme would have dedicated and consistent medical supervision and input. A larger cadre of nurses would spread the burden of the on-call duty but would dilute the valuable intimate knowledge of individual families and patients. Transition planning for transfer to adult services is difficult as there are no similar programmes available within the adult care settings. Finally, families who frequently present to ED should be electronically flagged for possible referral to the ACE programme. Similar to concerns at an open-access pediatric assessment unit for children with special needs,18 the ACE system potentially bypasses general practitioners, which could then reduce their experience in managing this group of children. However, this disadvantage is probably outweighed by ACE staff actively seeking support from general practitioners and the facilitation of information flow from specialty teams to general practitioners.

This study has a number of limitations. Although the study suggests positive outcomes it was non-controlled and undertaken at a single centre. Judgements by ACE nurses as to which ED presentations were unavoidable or which call would have led to an ED presentation or “avoided” one were subjective. We did not collect information on the demographic background of the enrolled families, which might be relevant for the analysis of care-seeking behaviour, nor did we collect data on other possible sources of care parents could have consulted resulting in an “avoided” ED visit. It is possible that over time the composition of patients enrolled in the ACE programme changed and that the changes in ED attendance rates occurred because of a different patient composition rather than changes in management. However, the baseline requirement of frequent ED attendance inpatients with long-term chronic complex problems did not change, and a review of enrolled patients in 2003 compared to 2006 showed a similar patient diagnosis and severity profile. The applicability of our findings to other institutions is limited where subsets of patients are cared for comprehensively by their subspecialty services. Data regarding waiting times, time of decision to admit and time of transfer are based on the accuracy of data entry by ED staff in the electronic ED log. Finally, direct data as to improved quality of care for ACE enrolled patients are difficult to come by.

CONCLUSIONS

Our intervention of an ED-based nurse-driven coordination programme for CSHCN who frequently attend the ED resulted in a moderate effect, probably an improvement in the quality of care rather than a quantitative change in management. It can enhance families’ capacities to manage their children’s conditions in the community. Key components of the ACE programme were 24-hour phone access, the development of electronically portable patient care plans, care coordination with complex care teams and a small group of experienced nurses who were able to build trust with families.

What is already known on this subject

  • Some children with complex health care needs frequently attend emergency departments.

  • There is a need to improve care for these children.

What this study adds

  • It is possible to create an coordination programme based in the emergency department for children with complex healthcare needs.

  • We were able to demonstrate a moderate effect of the intervention, probably an improvement of the quality of care rather than a quantitative change in emergency-department attendances.

Acknowledgments

We thank Prof. Jill Sewell, Centre for Community Child Health, and Assoc. Prof. Dinah Reddihough, Child Development and Rehabilitation, Royal Children’s Hospital, Melbourne for their thoughtful review of the manuscript.

REFERENCES

Footnotes

  • Competing interests: None.