Liberating the NHS: Transparency in Outcomes
Consultation Response from the
Royal College of Paediatrics and Child Health
Principles
1. Do you agree with the key principles which will underpin the development of
the NHS Outcomes Framework (page 10)?
Yes – largely. But, what is missing from this document is sufficient detail on the implementation of the framework. Stating that the Secretary of State will use the framework to hold the NHS to account at a national level is straightforward. However, what matters in practice is how the framework will be used at a more local level, for instance at the level of GP consortia. Regarding this, the document appears somewhat confused. On the one hand, it states that “It will not be used as a tool to performance manage providers of NHS care.” (Section 1.15) At the same time, though, it also states that “the Department of Health will support the development of metrics that allow intra-UK comparisons to be made.” (Section 2.23) It remains unclear, then, how the framework will be used at the level of consortia – if at all. If it not used at that level, but merely at the national level, then we would suggest it will make no practical difference to how the NHS operates. It will report practice, but not influence it. If it is used at a local level, then adjustments for local circumstances / case-mix will be required. This is a highly complex task, and it is for precisely this reason that in many aspects of NHS practice, proxy process and structural metrics are more appropriate (see next point).
2. Are there any other principles which should be considered?
We believe that greater attention should be paid to the importance of proxy process and structure measures where no suitable outcome measures are available. The consultation document acknowledges that at a local level these may be appropriate. However, we would argue that for some services, even at a national level, the most relevant measure is not an outcome but a structure or process.
Immunisation rates and provision of safeguarding services would be two such examples – one process, one structural. In neither case, even at the national level, are there any meaningful outcome measures.
3. How can we ensure that the NHS Outcomes Framework will deliver more
equitable outcomes and contribute to a reduction in health inequalities?
By making measures of equality (e.g. relative difference in mortality across socio-economic groups) part of the outcome framework to be used at both a national and local level. We believe that greater attention should have been paid to equality measures throughout this document.
4. How can we ensure that where outcomes require integrated care across the
NHS, public health and/or social care services, this happens?
By insisting on integrated commissioning for these services, and by developing pathway tariffs for them. We are concerned that that this framework was not developed in parallel with the public health outcomes framework and alongside whatever framework is used to measure the outcomes of social services and children’s services. For some years, the RCPCH has been calling for a greater degree of alignment in the goals of regulators, commissioners and providers. There is a significant risk of continuing such lack of coherence by isolating the NHS contribution to outcomes in the way this framework does. In particular, many of the health outcomes of children are necessarily dependant on a range of providers that cross sectors, and therefore it is essential that robust cross-sectorial outcome measures are developed for this cohort.
Five domains
5. Do you agree with the five domains that are proposed in Figure 1 (page 14) as
making up the NHS Outcomes Framework?
They are one acceptable way to stratify outcomes, though having said that if we focus too much on mortality, then given the overall low mortality rates in children, it is inevitable that children’s services will suffer as a result.
6. Do they appropriately cover the range of healthcare outcomes that the NHS is
responsible for delivering to patients?
One of the unintentional problems of creating these domains is that many children’s conditions would fit into all five categories. Cystic Fibrosis, for instance, is a long term condition with acute episodes of illness, where significant safety issues apply, and which frequently leads to premature death. This framework forces us to create five different outcome measures for this single condition, and if only one of these is kept for inclusion, it may not be the most appropriate one. There should have been some mechanism to enable outcome measures to be defined that cover the whole pathway of care. We are of course aware that NICE quality standards cover such pathways, but we are concerned about those conditions that because they are relatively low volume are unlikely to become of the 150 quality standards. Cystic Fibrosis would be an example.
Structure
7. Does the proposed structure of the NHS Outcomes Framework under each
domain seem sensible?
There is an admirable logic to the way the framework is constructed. However, we are concerned that the basket of indicators may not be sufficiently comprehensive. If this is correct, there is a danger that commissioners will focus on certain high volume, high cost conditions to the relative disregard of others. Children’s services are likely to do poorly in such a situation as in terms of cost their health services do not comprise a sufficiently large amount of expenditure. Such an approach, though, would be to the detriment of the long term health of our population. Children’s ill-health stores up a legacy that lasts a lifetime. Therefore, in the long run it makes sense to invest heavily in children’s health now, even if the benefits will not be seen for some time.
Domain 1 - Preventing people from dying prematurely
8. Is ‘mortality amenable to healthcare’ an appropriate overarching outcome
indicator to use for this domain? Are there any others that should be
considered?
The measure as it stands takes no account of years of life lost, and some adjustment to this should be considered.
9. Do you think the method proposed at paras 3.7-3.9 (page 20) is an appropriate
way to select improvement areas in this domain?
If we are going to focus on certain conditions, then this methodology is acceptable. However, by only focussing on certain conditions, it is inevitable that other conditions will receive relatively little attention. A more comprehensive approach is required that takes into account all causes of mortality in an appropriately equitable manner.
10. Does the NHS Outcomes Framework take sufficient account of avoidable
mortality in older people as proposed in para 3.11 (page 21)?
Possibly. But it does not take sufficient account of mortality in childhood, as there is no adjustment for years of life lost.
11. If not, what would be a suitable outcome indicator to address this issue?
12. Are either of the suggestions at para 3.13 (pages 21) appropriate areas of focus
for mortality in children? Should anything else be considered?
As well as adjusting all mortality for years of life lost, we would propose including:
· perinatal / infant mortality rates
· all respiratory deaths, 0-18 years
· deaths associated with chemotherapy
· survival outcomes for acute leukaemias
· number of adolescents dying from suicide that have had contact with NHS in previous three months
· deaths from meningitis
· traumatic brain injury - post-admission mortality.
Domain 2 - Enhancing the quality of life for people with long-term conditions
13. Are either of the suggestions at para 3.19 (page 24) appropriate overarching
outcome indicators for this domain? Are there any other outcome indicators
that should be considered?
Neither of the proposed indicators adequately captures outcomes for children or young people (see our proposal in question 14 below).
14. Would indicators such as those suggested at para 3.20 (page 24) be good
measures of NHS progress in this domain? Is it feasible to develop and
implement them? Are there any other indicators that should be considered for
the future?
As possible overarching indicators, we would propose:
· Child-specific PROM for basket of conditions (Asthma, Epilepsy, Cystic Fibrosis, neurodisabilities, diabetes, rheumatic conditions, Attention Deficit Hyperactivity Disorder, Autism, Inflammatory Bowel Disease)
· School absence due to the condition for basket of conditions (Asthma, Epilepsy, cystic fibrosis, neurodisabilities, diabetes, rheumatic conditions, ADHD, Autism, Inflammatory Bower Disease)
15. As well as developing Quality Standards for specific long-term conditions, are
there any cross-cutting topics relevant to long-term conditions that should be
considered?
An absolute essential for this is child protection. There exists no single, suitable outcome measure to assess safeguarding services and the only appropriate means by which to ensure high quality safeguarding services is a NICE quality standard that specifies the levels of service required across the whole pathway.
The proportion of children with long term conditions with multi-disciplinary care plans should also be considered as a cross-cutting measure.
The quality of ‘transitional care’ from adolescent to adult services should also be considered as should the development of self-management skills in adolescents (i.e. their ability to manage their own condition).
Domain 3 - Helping people to recover from episodes of ill health or following injury
16. Are the suggestions at para 3.28 (page 27) appropriate overarching outcome
indicators for this domain? Are there any other indicators that should be
considered?
· Unplanned returns to urgent / emergency care within 48 hours of being seen in urgent / emergency care – children only.
We believe that this indicator better captures the scenario, common in paediatrics, where a child is seen in urgent / emergency care, but the severity of their illness is not recognised, and so they are not admitted, only to return later and be admitted in a more unwell state.
17. What overarching outcome indicators could be developed for this domain in
the longer term?
18. Is the proposal at paras 3.30-3.33 (page 28-29) a suitable approach for
selecting some improvement areas for this domain? Would another method be
appropriate?
This is an appropriate methodology if the intention is to collect outcomes for the most common acute conditions requiring hospital admission. However, it is not an effective methodology to assess the overall quality of care in respect of acute conditions in childhood. As a measure of primary and secondary care, we would propose the indicator given above – namely:
· Unplanned returns to urgent / emergency care within 48 hours of being seen in urgent / emergency care – children only.
As a measure of in-hospital care, we would propose:
· Number of cardiac arrests, respiratory arrests and acute life threatening events for children that have been in the hospital more than 1 hour (excluding those taking place in Paediatric Intensive Care / Neonatal Intensive Care)
19. What might suitable outcome indicators be in these areas?
· emergency hospital admissions for children with gastroenteritis
· emergency hospital admissions for children with lower respiratory tract infections
· number of admissions of children in diabetic keto-acidosis
Domain 4 - Ensuring people have a positive experience of care
20. Do you agree with the proposed interim option for an overarching outcome
indicator set out at para 3.43 (page 32)?
No. The current surveys take no account of children’s experience of care (or their parents) and a new survey must be developed to address this. This is an area in which we have considerable experience. It is important to appreciate that the use of PREMs with children and young people and their parents / carers is relatively underdeveloped compared to its use in adult healthcare.
In our own work on feverish illness in children, we have already used questionnaires with parents / carers which explored their experience of urgent care provision. We are continuing this work by developing a PREM which will explore parents / carers and children's experience of acute illness in any urgent or emergency care setting, and additional ones which will address children’s experience of epilepsy and diabetes management. The intention is to make these PREMs available nationally once the pilot phases have been completed. Given that acute illnesses represent a significant proportion of all childhood illness, there is no reason why the urgent / emergency care PREM could not be utilised by the DH in domain 4 of its outcome framework. It could certainly feature, for instance, in the emergency care dashboard that is currently being developed.
It is also worth noting that BLISS have developed a questionnaire for parents of babies on neonatal units and it may be the case that this could be used in assessing their experience of patient care.
Clearly, further work is required to develop appropriate PREMs that can be used by children and young people themselves. However, our extensive experience of working with children and young people in clinical settings, and through our engagement with child health advocates in our youth advisory panel, the RCPCH is in a unique position to advise on this work.
21. Do you agree with the proposed long term approach for the development of an
overarching outcome indicator set out at para 3.44 (page 32-33)?
Possibly – as long as child and parent / carer specific questionnaires are developed, and contribute to the overall indicator. It is important to note that such questionnaires will need to be stratified according to the age group of the children and young people. Adolescents will require a different approach to those under 12 for instance. An important aspect of this is how we will reach and so gain the views of the non-respondents. This is especially important for young people as those who don’t initially respond are often those most in need of healthcare. It is also important that such questionnaires address the physical characteristics of care settings, and whether they are age appropriate. When children and young people have been surveyed on their experience of care, the things that bother them are often not those that concern clinicians e.g. the nature of the waiting area, the food available and how nice the receptionists were.
22. Do you agree with the proposed improvement areas and the reasons for