The NHS In Crisis: A Statistical Surgery
J.J. Patrick – Former Police Officer, Parliamentary Whistleblower on Crime Statistics, and Specialist Statistical Consultant.
The National Health Service is in a funding crisis, according to Chief Executive of NHS England, Simon Stevens, who appeared before the Public Accounts Committee and challenged the Government view the service had been allocated in excess of the operating funds it had requested.
Stevens’ appearance on the 11th of January 2017 was the latest in a short order of headlines declaring the service to be teetering on the brink of collapse, seemingly first sparked by the British Red Cross’ declaration of a ‘humanitarian crisis’ in the NHS on the 6th of January 2017, which appears to be supported by reports on the 10th of January 2017 of disastrous Accident and Emergency treatment times. Overnight, The Guardian reported that twenty hospitals have declared ‘black alerts’, which mean patient safety can no longer be assured.
The NHS is an institution of which we should rightly be supportive, it is there for every person from birth until death and the staff perform a phenomenal role in society and deserve high praise indeed. But what is the truth about this crisis in the NHS? Before beginning to unravel the answer to this question, it is necessary to provide some dispassionate clarification and comment on some of the headline issues which have dominated the last week.
The Operational Pressures Escalation Alert Levels Framework (the OPEL alert system cited by the Guardian) appears to be a relatively new policy, seemingly introduced under the NHS improvement scheme in October 2016. The full policy is a sensible, and simple, national standardisation framework to unite Accident and Emergency Department approaches across the country and ensure they are more closely tied with the REAP (Resource Escalation Action Plan) model in use across the Ambulance Service. The Framework also appears to aim for enhanced data collection and requires, since November 2016 according to documentation, a daily collation of SitRep data in a standard format, signed off by a duty director. It is, therefore, quite feasible that the OPEL system is currently being put through its first serious tour of duty and the use of the alert system is actually a healthy sign of its efficacy.
It is also quite clear that the policy is reliant on public communication of serious alert states, with it specifically outlining communication protocols to be used as a tool to drive demand reduction during peak periods: “It is recognised that at times of severe operational pressure, it may be necessary to communicate these pressures to the public to help manage demand and bring stability to the situation. Service disruptions are more likely to occur during winter, and when this happens there is a recognised need for local health and care leaders to communicate this via the press, to ensure local populations are well informed of pressures in their area.”
We can subsequently draw three rather simple conclusions from the ‘black alerts’ story:
1) 20 NHS Trusts have faced a peak period of demand, and selected the appropriate OPEL Alert Level which reflects the situation in the criteria set out in the new national policy, and have taken action as defined within the policy framework.
2) Those trusts have gone on to communicate the alert state to their communities in order to inform local people of alternative care options and to help reduce demand and reinstate a lower level of OPEL operating status. (As the Guardian article shows, the longevity of such high alert periods has been only hours, again indicating efficacy of the national framework).
3) The service disruptions and high grade OPEL alerts have arisen in a known and defined peak period for NHS service demand, which is specifically cited as requiring preparedness in the national framework as set above, and in the paragraph of the policy which immediately follows: “Local A&E Delivery Boards (and constituent member organisations) are therefore strongly encouraged to engage with local media ahead of winter to set out and explain the issues and processes to support effective communication with the public.”
The British Red Cross:
While the work of Charities is to be rightly and consistently commended, it is necessary to understand the context in which the British Red Cross are currently operating, in order to assess any identifiable rationale for their statement which may cast a different light upon their intervention. For the sake of informing the debate around the situation and their commentary it is necessary to refer to their current strategic vision and provide some background information on the current Chief Executive, Michael Adamson.
This article, from the publication Third Sector dated 2014, sets out a context in which the British Red Cross statement may be viewed more cynically: “Adamson had a range of jobs, including head of international development, during an initial 11-year spell at the British Red Cross. He then went to work in the NHS…He returned to the BRC in 2010. “There are more similarities between charities’ activities and those of organisations from other sectors than perhaps we acknowledge.””
While setting out a background of what is described as ‘enviable funding’, Adamson makes clear his wishes to expand the ‘BRC’ brand at home in the UK, particularly focusing on the provisions they make which tie in with NHS operations: “Despite its well-known name, Adamson thinks the BRC brand needs clarifying. The charity is associated largely with humanitarian work abroad, and Adamson says not enough people are aware of its wider work, such as first-aid training, supporting older people to live independently and helping to tackle food poverty here in the UK. “We want to be better known for all aspects of what we do,” he says. “It’s important for our influence on policy-makers, for funding and for the recruitment of volunteers. We want to get over that unique sense of our identity and that we’re not an NGO in the classic sense.”
The most relevant, and in regards of the current situation poignant, message Adamson delivers relates to obtaining a better financial recompense for the provision of services directly linked to the NHS: “When we look at our returns, there are some areas where we could recover more of the costs,” he says. “In our ambulance support work, for example, we have to get much better at charging appropriately to our partners to make sure we recover those costs.”
In the BRC’s statement of January 2017, in which the term ‘humanitarian crisis’ was used, they make specific reference to their home support and ambulance work, and call upon the government to “allocate immediate funding to stabilise the current system and set out plans towards creating a sustainable funding settlement for the future.”
We can draw three further conclusions from the ‘humanitarian crisis’ story:
1) There is no empirical data or other evidence which supports the definition of ‘humanitarian crisis’ and further, set against the OPEL Framework information above, it is likely that BRC have simply provided established charitable assistance during identified periods of peak demand.
2) BRC has a long term plan to increase ‘brand awareness’ of its home support, ambulance provision, and other UK health activities, and the reporting of this story has helped raise this profile. In addition, the Chief Executive has direct experience of working for an NHS Trust and has experience of the operational environment and demands.
3) BRC has a long term plan to pursue better financial settlements for its health related service provisions, seeing itself as more than a traditional NGO with sights set on policymaker influence. They have specifically called for immediate and long term public investment in line with these goals.
Is This Crisis Unprecedented?
The short answer is no. The NHS consistently suffers a winter crisis due to the so-called ‘flu season’ and has built winter needs into its demand framework, with a view to preemptive information campaigns forming part of the OPEL policy for demand management. Even restricting the summary to a recent three year period, the pattern is clear.
In November 2013, following numerous articles of an NHS funding crisis, The Guardian reported “Hospitals scramble to prevent crisis in NHS’s ‘toughest ever’ winter”, writing “Fears that the service may face its toughest ever winter have forced NHS trusts to use beds in nursing homes, reopen disused wards and build new ones to boost their capacity. At the same time trusts are recruiting nurses from abroad to tackle staff shortages so they can cope with the expected impact of flu, norovirus and bad weather…One trust has even obtained a prefabricated building to use as an overspill ward if it comes under particular pressure.”
In December 2014, the Express reported an “NHS beds crisis as flu hits three-year peak” stating emphatically “The NHS appeared on the verge of meltdown yesterday as it emerged hospitals are admitting record numbers of emergency patients.”
In January 2015, The Guardian again reported chaotic scenes across the country under the headline “‘Bedblockers’: the fit-to-leave patients deepening hospital crisis”, writing “Lack of care and home support means people are trapped in hospital beds needed for new patients, NHS bosses say”
Finally, in August 2016, The Telegraph reported figures which showed the “NHS in grip of worst bed-blocking crisis on record” citing extended waits in A&E, elongated ambulance response times, and “a near doubling in the numbers of patients stuck in hospital, for want of care at home, or help to get them discharged.”
On review of several years worth of reporting across a broad spectrum of the media, inexhaustively listed here for the sake of brevity, it is safe to conclude that the NHS has been showing effects of financial pressure and the need to accommodate increased demand, in particular around identifiable periods, for several years. It is also safe to say that the issues faced have been subject to little or no change over a prolonged period, and new management and improvement practices have only recently been implemented.
It is worth noting that the NHS owns some large degree of responsibility for the holding of beds due to some of its own working practices. For example, Pathology labs often do not operate full weekend cover, meaning that a patient could need to be held in a bed over weekend awaiting results for tests which could result in an earlier release. This is not a complex situation to resolve, yet it persists which raises some questions as regards efficiency and in some respects supports the Government’s views on the ‘7 Day Service’. A simple fix such as this could resolve a number of issues and should be taken forwards as a specific area of policy focus, to the benefit of patients and staff alike.
On the basis of a summary analysis, it cannot be said that events currently impacting NHS service delivery are unprecedented, nor unusual.
A Statistical Surgery:
NHS Data collection, though improving along with the policy framework, remains insufficient to produce real time public data for analysis. Subsequently it is necessary to extract available data from a series of sources to provide a round picture of funding and demand, in order to make some assessment of the true picture of the state of the health service. A caveat in terms of data integrity due to its nature, is that the data shall be assumed to have an accuracy threshold of >90% <95% due to factors of age and unknown factors in the quality of the raw data collection.
The Office For National Statistics:
Somewhat surprisingly the latest ONS statistical bulletin on UK Health Accounts is dated 2014. In summary, they state health care spending was £179.4 billion in 2014, a 4.2% increase on the 2013 spend, though as a percentage of GDP it remained static at 9.9%. During the same period, the government spend was £142.6 billion, which accounted for 79.5% of health care spending, an increase from 79.4% in 2013.
These figures clearly show an increase in total and governmental health care spending. It is worth noting that the UK’s GDP remained largely static during this period according to ONS charts, rising by a nominal less than 1%.
The NHS Confederation:
The NHS Confederation states it is the only membership body which brings together and speaks on behalf of all organisations that plan, commission and provide NHS services. The membership is drawn from every part of the health and care system. They provide a comprehensive summary of key statistics on the NHS, drawn from varying sources. As with almost all NHS data, the latest update was collated in November 2016.
In terms of the financial situation, NHS net expenditure (resource plus capital, minus depreciation) has increased from £75.822 billion in 2005/06 to £117.229 billion in 2015/16. Planned expenditure for 2016/17 is £120.611bn. In real terms the budget is expected to increase from £117.229bn in 2015/16 to £120.151bn by 2019/20. Health expenditure (medical services, health research, central and other health services) per capita in England has risen from £1,868 in 2010/11 to £2,057 in 2014/15. The NHS net deficit for the 2015/16 financial year was £1.851 billion (£599m underspend by commissioners and a £2.45bn deficit for trusts and foundation trusts).
These figures clearly show a significant increase in net expenditure over the last decade, and that planned expenditures far outstrip the current budget allocation – though this will almost be in balance if current expenditure is maintained, by 2020. The spend per capita has increased nominally, considering societal factors, which indicates a population change is driving total per capita costs to create higher expenditure overall. It’s clear there is a current deficit of a substantial nature, despite savings, and the NHS is commencing its next year in the overdraft with a commitment to overspend already made set to increase this deficit. It appears unlikely commissioner savings will be able to offset this.
It is worth noting, internationally, that the NHS was rated as the best system in terms of efficiency, effective care, safe care, coordinated care, patient-centred care and cost-related problems. It was also ranked second for equity. This rating was given by the Commonwealth Fund in 2014.
The NHS demand statistics are comprehensive and open with the headline fact that the health service deals with over 1 million patients every 36 hours. Some of the headline statistics relating to pathway waiting times have been excluded here as measures have since ceased in some cases.
In 2015/16 there were 40 per cent more operations (‘procedures and interventions’ as defined by Hospital Episode Statistics, excluding diagnostic testing) completed by the NHS compared to 2005/06, with an increase from 7.215m to 10.119m. There were 16.252m total hospital admissions in 2015/16, 28 per cent more than a decade earlier (12.679m).
These admissions and procedures statistics are valueless as they provide a decade wide comparison which gives no useful indication of current versus recent demand, and make no account for changes in data integrity and recording policy over time. The figures themselves present no surprising information and are largely reflective of a match in increased expenditure over the same period.
The total annual attendances at Accident & Emergency departments was 22.923m in 2015/16, 22 per cent higher than a decade earlier (18.759m). The proportion of patients seen within 4 hours at A&E departments in 2015/16 was 87.9 per cent in major (type 1 units) and 91.9 per cent overall. There were 3.140m category A calls (Red 1 and Red 2) that resulted in an emergency response in 2014/15, 9.3 per cent more than the previous year (2.872m). 71.9 per cent of Red 1 ambulance calls were responded to within eight minutes in 2014/15.
The decade comparison is, once again irrelevant and the target time data for A&E gives no indication of change in performance versus demand over time. The ambulance data, however, gives a clear indication of an increase in demand at 9.3% which should be directly reflected in A&E demand increases. With the introduction of walk in and 111 procedures this increase in emergency demand needs to be the subject of specific scrutiny to establish its drivers. The ambulance response time data is meaningless in this context, as it indicates nothing as regards changes in demand or performance.
The total number of outpatient attendances in 2014/15 was 85.632m, an increase of 4.4 per cent on the previous year (82.060m). In the 2015 calendar year, 482,120 NHS patients were admitted to independent providers for their elective inpatient care. There were 802,949 referrals made by GPs to independent providers for outpatient care during the same period.
These figures show a clear increase of 4.4% in outpatient attendances, with a large number of referrals made to outsourcing providers, with only just over half of those being accepted. It is impossible to identify trends in GP referral practice but the providers themselves should be scrutinised to ensure they are operating at a capacity capable of reducing demand on the NHS as appropriately specified within their contracts.
At the end of September 2016, there were 3.703 million patients on the waiting list for treatment. 348,542 (9.4 per cent) had been waiting for longer than 18 weeks, compared to 247,388 (7.5 per cent) at the same point in 2015. The number of patients waiting longer than a year for treatment declined from 20,097 in September 2011 to 214 in November 2013, before increasing again. In September 2016 the number stood at 1,181. Over the past three years, the number waiting in excess of 26 weeks has increased from 48,769 to 108,459 in September 2016.
The headline comparison of patients waiting is not comparable as the overall increase/decrease in demand is not accounted for. These figures are subsequently meaningless, even though they – on the face of it – indicate an increase in the number of patients waiting. While the percentage increase in patients waiting more than a year is substantial, the broader concern would be how these reductions were achieved, to a 1/20th of the figures recorded in 2011. Rather than an indication of short term demand increase, the substantial shifts are so significant they only indicate statistical deviance or ‘system gaming’.
Overall, it is possible to conclude that this data is incoherent and incomplete, and while there are indications of increasing demand, it is impossible to quantify. It is also clear that financial pressures do exist and that there is a need to bring the planned expenditure down through efficiency, or for the government to meet its 2020 spending target in the next twelve months. It is also clear that there are issues in the referral to outsourced providers system which urgently need to be explored. Further, it is clear that a comprehensive analysis of patients waiting for treatment needs to be carried out and reported upon.
Again some months out of date, NHS England provides a monthly statistical bulletin. The latest edition dated October 2016 does however provide a useful snapshot of near current NHS demand. Oddly, the reports appear to be released some six weeks after completion, with his having been published in December 2016. In summary, it states: “The long-term trend is one of greater volumes of both urgent and emergency care and elective activity, with A&E attendances up 4.5%, emergency admissions up 3.4%, diagnostic tests up 5.1% and consultant-led treatment up 4.3%. In the case of urgent and emergency care in October 2016, the NHS constitution standards were not met for A&E waiting times.”
In October 2016, there were 1,235,035 calls offered to the NHS 111 service in England, a 14% increase on the 1,083,628 in October 2015. Of calls triaged, 13% had ambulances dispatched, 9% were recommended to A&E.
These statistics show a clear increase in calls made to 111 of 14% over the course of a year. While this demand increase itself is useful, the figures which resulted in Ambulance calls or A&E directions hold no value as they aren’t comparable to previous figures. It is impossible to tell if 111 is a contributory factor to any increase in demand on emergency services and departments.
In England, since June 2016, only eight of the eleven Ambulance Services, covering 70% of the population, still use the Red 1 and Red 2 classification. Due to absence of commonality the ambulance response statistics are rendered meaningless and provide no clear picture of demand.
There were 2,000,645 attendances at A&E in October 2016, 4.0% more than in October 2015. Attendances over the latest twelve months are higher than levels in the preceding twelve month period (an increase of 4.5%). 89.0% of patients were admitted, transferred or discharged from A&E within four hours of arrival, below the 95% standard.
These figures show a clear, albeit low, increase in general demand on A&E departments over the 12 months, and the four hour service time is proportionally impacted by this increase.
Overall, it is impossible to define from the available figures whether the ambulance service has seen an increase in demand, and due to the data framework no meaningful conclusions can be drawn. In respect of A&E it is clear that there has been an increase in demand, directly relevant to an increase in treatment times. 111 indicates an increase in demand, though there is no way of identifying if this a simple diversion of demand away from the 999 system because the figures are not recorded.
By way of further exploration, NHS England also provide monthly data (from November 2016), on critical care bed availability. The data is quite comprehensive and allows a year on year, period on period comparison over time. The availability of critical care beds has clear implications on A&E departments as bed availability directly impacts the churn rates and generates trolley waits and emergency bed blocking. The report gives a summary that “There were 4,055 adult critical care beds available and 3,403 occupied, giving an occupancy rate of 83.9%. This is higher than the occupancy rate observed last month, which was 81.8%, and higher than in November 2015 (83.8%). There were 483 paediatric critical care beds available and 427 occupied, giving an occupancy rate of 88.4%. This is higher than the occupancy rate observed last month, which was 76.4%, and higher than in November 2015 (88.1%). The occupancy rate of 88.4% is the highest recorded since data was first collected in August 2010. There were 1,401 neo-natal critical care beds available and 1,018 occupied, giving an occupancy rate of 72.7%. This is higher than the occupancy rate observed last month, which was 70.5% and higher than in November 2015 (70.6%).”
With these figures clarity is provided that critical care demand has increased over the short term, and over the longer term, in the case of paediatric care resulting in the highest occupancy rate since data collection began. Overall, however, the increased demand levels are reflective of the increased demand on front line A&E departments, almost being logically proportional. However, reviewing the data over time, directly comparing demand periods from 2014 to 2016, it is clear that the number of available beds has increased, and demand for those beds month by month has remained largely static. In short, there is no apparent surge in critical bed demand which can be identified as outside of the expected statistical norms in any given area of care or time period. Simply put, NHS demand simply appears to be growing with the population and adverse effects are not made out in the statistics.
What is the truth about this crisis in the NHS?
Winter demand is a documented and well understood part of the NHS demand profiling, is fully accounted for in national policy, and has been well known for several years, with repeats of identical patterns every year. It cannot be said this is a new, unprecedented, or otherwise unique issue. In fact, the use of the OPEL system in line with the new national policy framework appears to show that the NHS is exceptional at preparing for and dealing with periodic demand peaks, though the communication of this needs to be adjusted. More work is evidently needed on demand mapping to ensure resource deployments are sufficiently intelligent to reduce known burdens in a more effective fashion.
While data integrity is an unknown quantity and much of the summary data collected is meaningless, what is available and serviceable indicates there is an increasing level of demand on NHS services, in particular A&E departments – though, at the documented levels, it cannot be said demand has changed so significantly as to warrant the application of a crisis status. Emergency response is a pinch-point in any service, and there are basic steps which the NHS has yet to take to alleviate some of this burden. In fact, the growth of critical bed provision and relatively static nature of occupancy rates over time indicate that the NHS is expanding provisions at a rate sufficient to manage its demand effectively.
Financially, the NHS has planned expenditure which exceeds its budget and is going to have to find savings – as it has shown it is capable of – but the Government is going to have make a reality concession and pay off the current overdraft if this plan is to stand any chance. As a world renowned service praised for its quality, which delivers per capita health care for exceptional value, Parliament is simply going to have to swallow this pill for the benefit of everyone. The financial contribution required is, in terms of the overall national budget, insignificant.
It is not possible to conclude the NHS is in crisis. It is, simply put, a complex public sector organisation which needs a nominal budget increase, more coherent national policy along the lines of the OPEL framework, more detailed, reliable, and current statistical data, and a drive to remove self-induced blockages which increase workloads at service pinch-points. There are a number of areas which require external scrutiny, and these should be pursued with specific focus by the select committees.
The interventions of third sector parties with vested financial and strategic interests are unhelpful to the NHS and the debate around it, and have in fact facilitated a large degree of ‘issue clouding’ which has resulted in ‘sensationalist’ reporting of what is, on the basis of carefully assessed evidence, the status quo.