Response rates for individual Delphi rounds were variable, ranging from 69% in R1 to 100% in R2 and 92% in R3, with 46% completing all three rounds. Although the response rate for R1 did not achieve Sumison’s27 70% standard, non-responders were spread amongst the subgroups, thus minimizing potential bias. Given that R1 was used to inform the development of the closed questionnaire used in R2 and R3, the reporting of the findings in this paper focuses on the quantitative analysis of these questionnaire rounds, in particular on changes between the rounds.
Integration and Coordination of Services
Responses to statements regarding integration and coordination of current services (including palliative care) are presented in Table 1, which includes findings from both rounds, allowing for changes and response stability to be compared. Wilcoxon’s Signed Rank Tests are also reported, with p ≤ 0.05 set as the significance level. As this section of the questionnaire asked experts to rate their agreement with the statements the overall level of agreement at R3 is also reported (reflecting the combined frequency of agree/strongly agree responses).
Table 1 about here
Examining statements relating to the integration and coordination of services (items 1 – 6 Table 1), expert agreement and consensus was reached on only two statements, and only group consensus on a third. Wilcoxon’s Signed Rank Test indicated stability of responses between R2 and R3 for all items. The group agreed that there are some difficulties relative to the interface between acute and community-based services, specifically that there is poor communication between the two, and that acute services are not aware of the range of services available to children and their families in the community. Two other statements approached, but did not achieve, the threshold of agreement. Seventy-five percent of the panel (n=9) agreed that there is poor coordination and integration of services, and the same proportion (75%) agreed that General Practitioners lack the experience and expertise required to care for children with LLNDD.
Within the context of integrated care, propositional statements related to the provision of palliative care to young children with LLNDD and their families achieved the lowest level of agreement in this study. Of the six statements, only one achieved both agreement and consensus in the final round (Table 1). Wilcoxon Signed Ranks Test indicated stability of responses with no statistically significant change to scores between R2 and R3. Specifically, the expert panel agreed that in palliative care is only considered late in the child’s condition or in cases of crisis management. The IQR for four other items was reduced between R2 and R3, indicating a move towards consensus, but the IQR for the statement “medical staff are reluctant to discuss the fact that children are life-limited with parents” increased from 1.75 in R2, to 2.75 in R3. One other statement approached the study’s agreement threshold, with 75% of the group (n=9) disagreeing that access to palliative care is readily available if it is required.
Group opinion on the approach adopted by health professionals was divided, although generally the trend was towards a negative perception, which was indicated by comparing the percentage of the group who agreed with the percentage who disagreed excluding the percentage of “not sure”. Fifty percent of the expert panel (n=6) agreed or strongly agreed that children often undergo futile interventions compared with 17% who disagreed or strongly disagreed (n=2). Sixty seven percent of the group (n=8) agreed that medical teams fail to recognize the palliative care needs of the child, with the same proportion agreeing that there is reluctance on the part of medical teams to discuss with parents the fact that the child’s condition is life-limiting . Finally, sixty seven percent of the group (n=8) disagree that there is a lack of interest in these children because of their limited prognosis.
Goals of Care
The goals of care are presented in Table 2 which includes the findings from both rounds allowing for changes and response stability to be compared. Wilcoxon’s Signed Rank Tests are also reported, with p ≤ 0.05 set as the significance level.
Table 2 about here
At R3, there was relatively little change to the rank order assigned to goals of care from R2. Five goals retained their priority status, while an additional seven moved up or down one ranking in R3. The priority of the goal “The child is cared for at home” demonstrated the greatest shift between rounds falling from second place in R2 to fifth place in R3, although the Wilcoxon Signed Rank Test demonstrated that there was no significant difference in the mean scores for this goal between rounds (W = .00, p=1.00).
While there was some movement in terms of priority awarded, the five highest ranking goals from R2 retained their top five ranking in R3. Three of these five goals achieved consensus in R3; “achievement of the best possible quality of life for the child” retained its number one priority; “open and honest communication with the family” increased from third priority in R2 second priority in R3; and “optimum symptom management” increased from fifth priority to fourth priority in R3.
Two of the top five goals of care identified by the expert panel failed to reach consensus in the final round. Although ranked as the third highest priority, “achievement of the child’s full potential within the limits of the illness” did not achieve consensus (IQR = 3). Three panelists provided commentary. For two the goal was synonymous with access to appropriate education and was inextricably linked with the issue of overall quality-of-life for the child. Similarly, although identified as the fifth ranked goal, “the child is cared for at home” did not achieve consensus in this round (IQR = 3.5). Exploration of the commentaries provided suggests that while in many cases this is an important goal, in others it is neither desirable nor achievable.
The lowest ranking goals “the child’s life is prolonged” and “the family is provided with the hope that things will get better” retained their position between R2 and R3, and both achieved consensus in R3, suggesting that the expert panel agreed that these were not priority in the care of young children with LLNDDs and their families.
Overall, the IQR for eleven of the goals of care were reduced between R2 and R3, demonstrating a move towards consensus. This did not hold true however for two goals; “parents are supported with the provision of care” and “achievement of a seamless web of care”. These demonstrated a small increase in IRQ of .25 and 1.25 respectively. Wilcoxon’s Signed Rank Test indicated stability of responses between R2 and R3 for 12 of the goals. The only goal not to achieve response stability was “the family continues to function as a unit and enjoy life” (W = -2.03, p = .04).
Priorities for Improving Services
Six change priorities for improving services retained their original ranking between R2 and R3 (Table 3), with an additional ten moving up or down one ranking between the rounds. Wilcoxon Signed Rank Test demonstrated that there was no significant difference in the mean scores for many of the priorities for improving services between rounds; with significant values reported for a single care plan for use across all services (W = -2.23, p=0.03) and a national directory of services (W = -2.04, p = 0.04).
Insert Table 3 here
While there was some movement in terms of priority awarded, the five highest ranking priorities for change from R2 were retained as the five highest ranking priorities in round three. Three of these achieved consensus in the final round. These included “a single care plan for use across all services” ranked as the highest priority change, “a greater level of communication between all the health professionals involved in the care of the child” ranked second highest priority, and “a key worker available to every family” ranked third.
Although ranked fourth and fifth respectively, neither “a greater level of coordination and integration of services” nor “a specialist pediatric palliative care consultant to act as a resource when necessary” achieved consensus in R3 with IQRs of 3.5 and 6.25 respectively, indicating a wider range of disagreement amongst the group with regards to the final ranking of these priorities. The IRQ was reduced for the remaining 13 items in this subscale in R3, suggesting a move towards consensus in this round. However, overall the range of IQR remained wide (3.5 – 9.5) which suggests that, excluding those items which were ranked as being the three most important priority changes to services, and on which consensus was reached, there was relatively little agreement amongst the panel as to the priority service changes should take to improve the care provided to children with LLNDD and their family.