Dr. Marie-Claude Grégoire
Featured article: Ringuier B, Troussier F, Boussicault G, Chapotte C, Rachieru P. [Non invasive ventilation and pediatric palliative care. A French survey]. Arch Pediatr. 2017;24(8):712–719. doi: 10.1016/j.arcped.2017.05.007. Epub 2017 Jun 28.
As I looked at the list of articles for the month of September 2017, I realized that not knowing which article to pick among an abundance of interesting articles isn’t necessarily a bad thing. The study by Ringuier et al discusses the use of non-invasive ventilation (NIV) in pediatric palliative care in France, and it caught my attention for many reasons. First, it is a topic that comes up regularly when discussing the long term care of many patients in my clinical practice as a pediatric palliative care physician. In discussions with colleagues from across Canada, I know that our situation is not unique. Second, the paper was published in French. There are many high-quality articles published in languages other than English that too often get missed, so I readily took the opportunity to highlight this one. I will summarize the main findings in lieu of official translation.
The study surveyed 186 clinicians in metropolitan France: pediatric respirologists, palliative care physicians and pediatric intensivists. The main question was: in a situation of acute or chronic respiratory distress, would you consider it “reasonable” to use non-invasive ventilation (NIV) in two groups of children with different goals of care. The first group included children with a “do not intubate” order, and the second group was made up of children with a comfort care only order. Information about the characteristics of the NIV prescription was also collected. The overall response rate was 34%, which is in the usual range for this type of study, although the response rate for the palliative care physicians group was lower (22%).
A majority of specialists surveyed found using NIV reasonable when treating children with a “do not intubate” orders, in both acute and chronic respiratory insufficiency cases. Not surprisingly, a smaller percentage, only about a third of respondents, found it reasonable to use NIV in children with a “comfort measure only” order. The specialists’ responses were very similar when asked about their prescribing experience in similar scenarios with one exception: non-intensivists had prescribed NIV for children with a “do not intubate” order with acute respiratory insufficiency and intensivists, even if both groups considered reasonable the use of NIV. In terms of other treatments offered, oxygen was selected as appropriate for almost all children while morphine and sedating medications were selected more frequently for children with a “comfort measures only” order. As for symptoms, breathlessness was the symptom cited the most frequently as indication for NIV for both groups of children. Of all symptoms listed, only two differed between the two groups: daytime hypercapnia and weight loss were selected more frequently as an indication for NIV in the group with a “do not intubate” order vs. the group with “comfort measures only” order. Finally, the criteria used to assess the efficacy of the NIV were similar in both groups in terms of the increase in comfort; the child and parental satisfaction; the decrease time in hospital; and the diminution of dyspnea the most important ones.
Despite some limitations (for example, not having parents or patients participate in the study) Ringuier et al collected important information that may be generalizable outside of France. The fact that breathlessness was used as the most important indicator for both groups demonstrates that comfort was a key factor in the decision to initiate NIV, in line with the common palliative care approach. This point is an important one for the population of patients we often work with in pediatric palliative care, especially in tertiary care centres. In our experience, many families are looking for a compromise between no ventilation in any form and mechanical ventilation provided in an invasive way (endotracheal or tracheostomy tube), for both acute and chronic respiratory insufficiency. Comfort is often the main family goal, and for some families NIV can be seen as the most comfortable option. The relatively wide range of clinician’s responses in the study may be partially explained by the fact that “comfort”, like “quality of life” is very subjective, multifactorial and individual-dependant. The lack of pediatric dyspnea assessment scales, especially for children who cannot self-report, further complicates the overall comfort assessment of children. Another good reminder from this paper is about the importance of having criteria, ahead of the NIV trial, to decide if the intervention is helpful or not. Too often interventions are started first and outcome measures established second, which changes the level of decision being made from an ethical point of view.
In conclusion, this paper confirmed what clinical experience has already taught us over the years: each intervention, like NIV, can be chosen or not chosen by families or clinicians using the same criterion of comfort. When it comes to establishing goals of care and medical interventions, guidelines help, but open discussion with families are crucial to understanding the multifactorial environment in which the child lives.
Dr. Bill Splinter
Featured articles: Quinn C, McCarthy S, Devins M, O’Reilly M, Twomey M, Ling J. Prioritisation of future research topics in paediatric palliative care in Ireland: a Delphi study. International Journal of Palliative Nursing. 2017;23(2):88–97. DOI: 10.12968/ijpn.2017.23.2.88
Bender HU, Riester MB, Borasio GD, Fuhrer M. “Let’s bring her home first.” Patient Characteristics and Place of Death in Specialized Pediatric Palliative Home Care. J Pain Symptom Manage. 2017;7(17):30212–9. DOI: 10.1016/j.jpainsymman.2017.04.006
Again, we have another month of many very interesting papers in the August list of citations. Two of the papers will be discussed in more detail, and they are markedly different.
The initial article by Quinn et al reports the results of their Delphi Study of the research priorities in children’s palliative care in Ireland. The study is limited by their focus on a single nation, but much of their results can be readily adapted to other countries. Anyone contemplating research and/or a quality improvement initiative within the highly specialized field of pediatric palliative care would be well advised to review this article in detail.
Quinn identified 8 research priorities and these include good clinical governance, national pediatric palliative care strategy, training programmes, children’s rights, linkages between maternity services and children’s palliative care, needs of the family caregivers, database, care planning and bereavement. All of these priorities were considered ‘extremely important’ or ‘very important.’ Quinn provided an excellent review of a Delphi study and they also discussed the five themes identified by their process.
The second article by Bender et al reviews the Munich experience of caring for 212 consecutive pediatric palliative care patients from 2009-2015. Of their study population, only 55% had died, and among that 55%, 84% died at home, 12% at the hospital and 4% at the hospice. These are very interesting numbers and atypical when compared to Canadian norms. In our setting, with a hospice adjacent to the hospital, almost all of our families opt for hospice care at the end of life. Our hospice functions quite independently from the hospital but there is marked synergies, such as we can readily transfer patients to and from the hospital and hospice. Also, most families develop very close ties with hospice staff before the end of life course, thus feel very comfortable having end-of-life care at the hospice as opposed to their home or the hospital.
In addition to the characteristics of patients receiving care under the specialized pediatric palliative home care (SPPHC) framework, it would have been nice if the investigators reported all of the local paediatric deaths (i.e. are all deaths referred to SPPHC or just some, and if so, what are the descriptors for the patients who are not cared for by the SPPHC?). That said, they do report national numbers as comparators. The age 0-1 yr are quite dissimilar (i.e. SPPHC population is markedly different than the national numbers). Bender et al report that those in their 1 to 20-year-old age group are similar (i.e. SPPHC numbers are similar to national values, with respect to their diagnoses). But upon my review of their Table 2, I certainly do not get the same impression; for example the national average is 30% of the children that die, it is due to a neoplasm while their local, SPPHC population has a 51% incidence of neoplasm.
Overall this is a very interesting article and it is very helpful to see what is the typical practice in Munich, Germany. This facilitates comparison between countries and localities. Of particular note, is their request for using common approaches to patient categories/classifications. They do recommend the use of ICD-10 as the classification of choice for children within the paediatric palliative care population. I would tend to concur with their recommendations.
Dr. Hal Siden
This month brings several good citations, many of them abstracts from the Royal College of Paediatrics and Child Health (UK) annual conference. In this month’s review I am going to highlight 2 full-length research articles I found of interest and comment on a third one.
Two articles worth reading because they are well-done studies and have implications for practice:
The 3rd article is: Jarvis, S., et al. (2016). “How many children and young people with life-limiting conditions are clinically unstable? A national data linkage study.” Arch Dis Child. Published Online First: 28 September 2016 doi:10.1136/archdischild-2016-310800
For 20+ years we have been discussing and debating the numbers of children who might need PPC services. This began with an estimate made in the late 1980s based on weak data, and we have contended since then with how to count prevalence of children “at-risk” of dying from a disease-related cause. Work by Chris Feudtner, Richard Hain, Simon Lenton, and others with great support from the charity Together for Short Lives has moved this work forward.
Dr. Stuart Jarvis, Dr. Lorna Fraser, and colleagues from York University, Martin House Hospice, and the Children’s Hospice Association of Scotland (CHAS) use a sophisticated methodology to determine how many children [in Scotland] are ‘clinically unstable’ and therefore eligible /appropriate for pediatric palliative care services. Jarvis’s report is a secondary analysis from a larger study of children’s palliative care needs in Scotland. The overall estimate in that report is very high: 95.7/10,000 (and even higher for infants). Behind the research was previous work by Fraser et al to refine ICD-10 codes with validation against the linked databases and pediatric palliative care service records, thus generating a useful list of conditions where a shortened life is likely.
Jarvis’s secondary report focuses on those who are defined as unstable, deteriorating, or dying. The authors operationalized a concept recently developed as a model by Hughes-Hallet for the UK National Health Service. They used database linkage, including hospital and community encounter records, to determine the number of children going into hospital, ICU and dying. The clinically non-stable group is estimated at 13.7/10,000. This number compares favorably with estimates of children in need of PPC services derived from other studies, including the most recent Together for Short Lives report (16/10,000) and work we did in British Columbia (9.8-15.3/10,000).
We need to continue to work in this area; the estimates are getting better and suggest that we focus on the estimates for the non-stable group. Comparing those estimates to actual numbers of children followed at centres in the UK and Canada suggest that either the estimates remain too high, or we have a major referral gap. This needs to be an subject of active inquiry. There will never be a final “right” number, but within similar settings (e.g., highly industrialized settings), estimates should begin to coalesce. Getting agreed-upon estimates is important for program leaders who need to plan for service delivery and deliver messages to funders such as donors and governments.