Our initial findings - 2020 Delivery



Deaths in the UK have risen by more than 40,000 per annum since 2011. Why?Exploring the causes of increasing mortalitycenter1828802nd February 201740000200002nd February 2017lefttop00IntroductionFor more than 30 years, the number of people dying in the UK was falling. However, this trend changed in 2011. Since then, deaths in in the UK have risen: there were 50,000 more deaths in 2015 than in 2011, and more than 40,000 more deaths in 2016 than in 2011.? Figure SEQ Figure \* ARABIC 1: Long-term trend of births and deaths in the UK (ONS data)The increase in deaths has principally been amongst the very elderly, and has been large enough to begin to cause a reduction in life expectancy for people aged over 85.Figure SEQ Figure \* ARABIC 2: Long-term trend in UK Life Expectancy at age 85This change to the trend in life expectancy was not expected or predicted by ONS:Figure SEQ Figure \* ARABIC 3: Trend in UK Life Expectancy at age 85 - actual vs forecastThe effect has been large enough to halt the previous trend of year-on-year reductions in age-standardised mortality. Life expectancy at birth is still growing, but the rate of growth is now substantially slower than at any point in the last 35 years – for much of the 2000s, male life expectancy at birth was growing at a rate of more than 100 days per year. This rate has now slowed to fewer than 40 days per year:Figure SEQ Figure \* ARABIC 4: Long-term trend in annual rate of change in life expectancy at birthThis is major change to the trend of growing life expectancy that had become so well-established that growing life expectancy is taken for granted by policy makers and the general public. So far, there has been little awareness or public discussion about this new trend, although the increase in mortality has been noted in the Office of Budget Responsibility’s “Fiscal Sustainability” publication in January 2017. While some may say that the trends of falling deaths and rising life expectancy had to end at some point, it has happened sooner and faster than anyone had been forecasting. Furthermore, the UK’s life expectancy remains below that of peer group countries such as France, Spain and Italy. We would like to start to understand the detail of what is happening, what are the causes, and what are the implications for all of us. This document is a first step, using publicly available data throughout. In a situation such as this, it is rarely possible to be certain as to root causes (for example, there is still argument 25 years on as to why murder rates in US cities suddenly dropped in the early 1990s). It is however possible to disprove some potential causes and to point to more likely causes that are consistent with the evidence. In that context, it’s important that the evidence is assessed, and that the likely root causes of the increase are understood. Through this document we aim to get the discussion started and we encourage as many people as possible to contribute to our exploration of the evidence here.Our initial findingsThis document addresses the following questions which explore the nature of the trends:Is the increase in deaths statistically significant?Yes, relative to the previous ONS baseline which projected deaths not increasing before end 2016, the increase in deaths is statistically significant (above 99.5% significance), as shown in figures 5 and 6 in the appendix to this document. This means that we can be 99.5% confident (or more) that the increase is due to something more than random variation. Has the increase been sustained over time, or could it have had a short-lived and external cause, such as a flu outbreak or the failure of a flu vaccine?Yes, the increase has been sustained now for five years, with each of years 2012, 2013, 2014, 2015 and 2016 seeing more deaths than in 2011.Has the increase applied in all parts of the UK, or has it differed across the different constituent countries of the UK?Yes, each of the countries of the UK saw a growth in deaths between 2011 and 2015. As shown in figure 7, Scotland saw the lowest percentage increase (7.4%), Northern Ireland the highest (9.5%), with England (9.4%) and Wales (9.1%) in betweenIs the increase in deaths sufficient to have caused an increase in age-standardised mortality and an absolute increase in death rates (i.e., deaths per 1000 population per year) within particular age-group cohorts – and in particular amongst the elderly?Figures 8 - 13 in the appendix to this document show that:The previous trend of year-on-year reductions in age-standardised mortality halted in 2011 – since then age-standardised mortality has fluctuated, but not shown significant trends up or downThe increase in deaths since 2011 is predominantly caused by an increase in deaths amongst the over 85sAmongst over 85s, annual mortality rate per 1000 population has risen since 2011, after a previous trend of falling annual mortality rates amongst this groupAmongst those aged 75 – 84, the previous trend of declining annual mortality rate per 1000 population has flattened outAmongst those aged 65 – 74 and those aged 0 – 64, the previous trend of declining annual mortality rate per 1000 population has continued with only a very slight amount of flatteningIs the increase in deaths sufficient to have had an effect on life expectancy (looking in particular at life expectancy at age 85)? Yes, across the UK life expectancy at age 85 showed a 30-year trend of steady year-on-year growth between 1980 and 2010. As shown in figure 14 in the appendix to this document, since 2010 the trend has become flat for men, while life expectancy has fallen slightly for women.If we disaggregate changes in Life Expectancy at age 85 by country within the UK, all four countries have seen a slight increase in male life expectancy at age 85 between 2009-11 to 2013-15 (ranging from 0.2% to 0.7% total increase across the period); all four countries have seen a decrease in female life expectancy (ranging from a 0.2% decrease in Northern Ireland to a 2.2% decrease in Wales, with England and Scotland having 0.7% and 0.5% decreases respectively). Figures 16 and 17 show details of this.Is the increase in deaths sufficient to have had an effect on life expectancy at younger ages (looking at life expectancy at birth and life expectancy at age 65)? As shown in figures 18 - 21, life expectancy at birth and at age 65 are still increasing but the rate of increase has dropped very substantially – the rate of increase in life expectancy at these ages is now slower than at any point in the last 35 years. In the case of female life expectancy, it is now increasing at no more than 25% of the rate at which it was increasing during the 2000s.Are there particular disease groups that have been responsible for the increase in deaths? As shown in figure 22, the principal change in recent years has been the increase in number of deaths coded to dementia and Alzheimer’s (up from 25,000 per annum in 2010 to 62,000 per annum in 2015). It is not clear how much of the increase in deaths coded to dementia and Alzheimer’s has been due to changes in coding practice (the government has incentivised higher diagnosis rates for dementia) and how much is due to actual change in cause of death.Apart from dementia and Alzheimer’s, trends in cause of death by other disease types have been quite gentle: there has been a small reduction in deaths from circulatory diseases, a small increase in deaths from cancer, and a small increase in deaths from respiratory conditions. Has the change in trend in life expectancy been picked up by policy-makers, for instance in the Office of Budget Responsibility’s forecasts of fiscal sustainability?In the Office of Budget Responsibility’s “Fiscal Sustainability Report” of January 2017, which used data from the October 2015 update of the Office for National Statistics’ population projections, OBR makes the comment that “… the ONS projections assume higher mortality rates, which reduces spending by 0.1 per cent of GDP in 2065-66. The cumulative impact of lower spending would reduce public sector net debt by 3.0 per cent of GDP by the same date; … under our interpretation of the longevity link on the SPA, the downward revision to life expectancy due to higher mortality rates would push back the profile of SPA rises … “The ONS publication from October 2015 (which takes account of 2012 – 2014 data on higher deaths, but predates data on 2015 and 2016 increases in deaths), projects continued increases in life expectancy at birth, but projects a slower rate of increase than had been assumed in ONS’s 2012 population projection. This document addresses the following questions which explore the root causes of the trends:Was the increase in deaths caused purely by the combination of growth in UK’s population and ageing of the UK’s population (i.e., baby boomer generation becoming elderly)?No: as shown in figure 10 in the appendix to this document, mortality rate per 1000 has increased amongst the elderly. As shown in figure 23, ONS was previously predicting that the effect of the UK’s ageing and growing population would be that annual mortality would reach a minimum in 2016, and then grow gradually thereafter. It was predicted that annual deaths would take until 2025 to grow by 50,000 per annum – whereas in actual fact, this has happened already in 2015Was the increase in deaths an effect that has been seen across multiple comparator countries – indicating a common cause across many countries?No: the effect seen in the UK is quite different from the effects seen in, for example, Spain, France and Germany (shown in figures 24 - 26 in the appendix to this document). We note however that so far Eurostat data for deaths in Spain, France and Germany only goes up to 2014, whereas we have data for the UK up to 2016. These conclusions will need to be developed as Eurostat publishes data for 2015 and 2016.Was the increase caused by climatic effects, for instance by cold winters in the UK?No: mortality in the UK is seasonal, and is typically highest in winter when the UK sees a higher burden of respiratory disease. Winter mortality is typically highest in cold winters.As shown in figure 27, the UK has not experienced abnormally cold winters in recent years, and as shown in figures 5 and 6 the increase in mortality has been a sustained phenomenon year-round – it hasn’t been caused by cold weather in winters (and nor has there been a material effect from very hot weather in summers).Was the increase caused by a rise in “lifestyle diseases”, including obesity, amongst the very elderly?Uncertain/ potential effect: the extent of the contribution of increases in “lifestyle diseases” to the increased mortality since 2011 is not yet clear. As shown across figures 28 - 31, it is known that prevalence of “lifestyle diseases” such as obesity and diabetes have been rising over the last 30 years, including amongst elderly cohorts. The rise in prevalence of these conditions has caused increased health risk (including risk of mortality). However, the increase in obesity has flattened since 2010, and based on the public data assessed so far we do not yet have causal evidence to link these factors in a quantified way to the observed change in mortality since 2011.Was the increase caused by a change in environmental factors, potentially associated with diesel and/or air pollution, since 2011?Uncertain/ potential effect: as shown in figure 32, it is known that air pollution has regularly exceeded limits in a number of cities in the UK both before and during this period, and there is evidence that air pollution contributes to circa 40,000 deaths per annum in the UK. Furthermore, it is known that diesel traffic has increased over the last decade, increasing particulate pollution in certain areas. However, the publicly available data that we have explored so far does not show causal evidence to link these factors in a quantified way to the observed increase in mortality since 2011.Was the increase caused by a real terms reduction in social care expenditure since 2011?This is a credible potential root cause but the magnitude of the effect is not yet clear. There has been a real reduction in social care provision to older people in the UK since 2011. As shown in figures 33 - 35, social care expenditure has reduced in the UK since 2011, and the number of elderly people qualifying for (means-tested) social care support has also reduced over this time. Furthermore, the threshold of need at which local authorities provide funded social care support has increased. While a link between reduction in social care provision and increased mortality is credible, the public data that we have examined so far does not provide causal evidence to directly link these factors in a quantified way to the observed change in mortality.Was the increase caused by declining NHS performance (e.g., A&E waiting times; delayed transfers of care; referral to treatment times) on a number of key measures since 2011?This is a credible potential root cause but the magnitude of the effect is not yet clear. NHS performance has declined substantially across several indicators since 2011, including A&E waiting times, delayed transfers of care and referral to treatment times – as shown in figures 36 - 38. Furthermore, there is evidence that links crowding of hospitals (which happens when waiting times increase) to mortality, and evidence that links delays to mortality. While a link between NHS performance and increased mortality is credible, the public data that we have examined so far does not give causal evidence to directly link these factors in a quantified way to the observed change in mortality.Was the increase caused by declining mental health within the elderly population, for instance related to anxiety, depression and loneliness?This is a credible potential root cause but the magnitude of the effect is not yet clear. Diagnosed prevalence of depression (across all ages) is increasing, and diagnosis rates of depression are higher amongst the elderly than amongst other age groups (see figure 39). It is widely acknowledged that depression and loneliness are material problems for the elderly in the UK. Furthermore, we know that poor mental health is associated with worse physical health and lower life expectancy. However, the public data examined so far does not give causal evidence to link these factors in a quantified way to the observed increase in mortality.Was the increase caused by increasing inequality, and/or by increasing poverty within more deprived communities?The absolute poverty rate among pensioner households is at a historic low, a trend that is broadly mirrored across all households (figures 40, 41). This suggests that rising poverty is unlikely to be the cause of increasing mortality and falling life expectancy. However, the rate of poverty reduction has dropped in recent years, with little reduction in pensioner poverty since 2009-10. It could be that declining poverty has been a cause of rising life expectancy over much of the last twenty years but this driver has now been removed, so other drivers that reduce life expectancy are now predominant. Confirming this would require further investigation.The increase may have been caused by more than one of the factors above, acting together. It is likely that there may be other potential causes too, that we have not yet considered.Our appealWe believe that the increase in deaths is both real and important – and that if so it is important that the causes and implications of this are understood. We are appealing to others to collaborate with us on investigating and understanding the causes of this rise, and indeed to launch their own projects to investigate this further. We hope to stimulate a debate that will lead to greater understanding both of the implications and causes of this change in mortality rates.About 2020 Delivery2020 Delivery is a specialist public service consultancy. Our staff are committed to improving public services and delivering lasting change for service users and taxpayers. Further details about the company can be found at The authors of this document are Russell Cake (one of 2020 Delivery’s Directors, who can be contacted on russellcake@) and Phineas Hodson (Senior Consultant at 2020 Delivery).This version: 2nd February 2017Appendix: Detailed AnalysesExploring the increase: 1. Is the increase in deaths statistically significant?Yes – relative to a previous baseline projection from ONS that showed annual deaths not increasing before the end of 2016, the increase in deaths that we have seen since 2011 is statistically significant:2016 was the 5th year in succession to see higher deaths than 2011 (in 2016’s case, provisional data published on 24/01/17 indicates that the number of deaths was higher than in each of 2011, 2012, 2013, 2014, but approximately 1% lower than in 2015).During each of 2015 and 2016, there were long runs of weeks (sometimes up to 13 weeks in succession) where every week saw higher deaths than the average of the equivalent week for the last 5 years. The probability of there being a run of 13 weeks in succession with higher than average deaths, at some point during a 52 week period, and of that occurring purely by chance (without a causative driver) is less than 0.5%. The probability of there being two runs of 13 weeks in succession with higher than average deaths, at some point during a 52 week period, and of those two runs occurring purely by chance (without a causative driver) is much less than 0.1%. See figures 5 and 6 for detailsFigure SEQ Figure \* ARABIC 5: Weekly deaths in 2015 vs average of previous 5 yearsFigure SEQ Figure \* ARABIC 6: Weekly deaths in 2016 vs average of the last 5 years This means that we can be more than 99.9% confident that what we are seeing is not just a picture of flat underlying annual deaths, with random variation causing the increase. Instead, the increase has a “real” driver of some sort.Exploring the increase: 2. Has the increase been sustained over time?Yes: 2016 was the 5th year in succession to see higher deaths than 2011 (in 2016’s case, the number of deaths was higher than in each of 2011, 2012, 2013, and 2014, but approximately 1% lower than in 2015.)This makes it very unlikely that the cause would be a single flu outbreak (which wouldn’t last across several years), or that the cause would be the failure of a single flu vaccine (as happened during winter 2014-15).Figures 5 and 6 (above) also show how the increase has been sustained over time. Exploring the increase: 3. Has the increase applied in all parts of the UK, or has it differed across the different constituent countries of the UK?Yes: Figure 7 below shows the picture for England, Scotland, Wales and Northern Ireland. The country with the lowest growth in deaths between 2011 and 2015 was Scotland (which saw 7.3% growth in deaths). Figure SEQ Figure \* ARABIC 7: Medium-term trends in deaths in England, Scotland, Wales and Northern IrelandExploring the increase: 4. Is the increase in deaths sufficient to have caused an increase in age-standardised mortality and an absolute increase in death rates (i.e., deaths per 1000 population per year) within particular age-group cohorts – and in particular amongst the elderly?As Figure 8 (below) shows, the increase in deaths has been sufficient to halt the previous downward trend in age-standardised mortality: since 2011, age-standardised mortality has been approximately flat. Figure SEQ Figure \* ARABIC 8: Medium-term trend in age-standardised mortality rates in England and WalesAs figure 9 shows, the rise in deaths since 2011 has been driven by a rise in deaths amongst the over 85s:Figure SEQ Figure \* ARABIC 9: Trend in UK deaths by age band (source: ONS data)As figure 10 shows, the increase in deaths amongst over 85s has been faster than the growth in population of over 85s. As a result, mortality rates per 1000 population have been rising in the over 85s since 2011: they reached a minimum of 143 per 1000 in 2011, before rising to 155 per 1000 in 2015.Figure SEQ Figure \* ARABIC 10: Annual mortality rate per 1000 population, UK people aged 85 and overFor people aged 75 – 84, there has not been a significant increase in mortality rate per 1000 since 2011, but there has been a flattening of the previous downward trend, as shown in figure 11 below:Figure SEQ Figure \* ARABIC 11: Annual mortality rate per 1000 population, UK people aged 75 - 84For people aged 65 – 74, there has been much less of an effect – only a very slight flattening in the downward trend in mortality, as shown in figure 12 below:Figure SEQ Figure \* ARABIC 12: Annual mortality rate per 1000 population, UK people aged 65 - 74For people aged 0 - 64, there has also been very little in the way of a visible change since 2011 – only a very slight flattening in the downward trend in mortality, as shown in figure 13 below:Figure SEQ Figure \* ARABIC 13: Annual mortality rate per 1000 population, UK people aged 0 - 64 yearsExploring the increase: 5. Is the increase sufficient to have had an effect on life expectancy?Yes: Figure 14 below shows ONS data on UK Life Expectancy at age 85, which shows a 30-year trend of steady growth between approximately 1980 and approximately 2010. After 2010, the trend flattens out both for male life expectancy, and shows a slight fall in female life expectancy.Figure SEQ Figure \* ARABIC 14: Long-term trend in UK life expectancy at age 85Figure 15 (below) shows that the flattening of the trend for life expectancy (and, in the case of female life expectancy, the slight decline since 2011) is very different from what ONS was projecting in 2010.Figure SEQ Figure \* ARABIC 15: Actual vs projected Life Expectancy at age 85, UK-wide, ONS data and projectionsBased on figures 14 and 15, it appears no longer to be true that “life expectancy is rising amongst the very elderly” – for those age 85 or more, life expectancy is now falling for women, and is flat for men. Figures 16 and 17 show long-term trends in Life Expectancy at age 85, split by the constituent countries of the UK. These figures show that all four countries have shown a flattening trend after 2011. There are some differences between the countries:All four countries showed rapid growth in life expectancy throughout the 2000s, but England showed a faster growth (both in male and female life expectancy) than other other country of the UKAll four countries show similar trends since 2011 in male life expectancy at age 85 All four countries saw a peak in female life expectancy at age 85 in 2011. Since then, Wales has seen the largest decline; no country has seen an increase since 2011Figure SEQ Figure \* ARABIC 16: Medium-term trend in male life expectancy at age 85, by country of the UKFigure SEQ Figure \* ARABIC 17: Medium-term trend in female life expectancy at age 85, by country of the UKExploring the increase: 6. Is the increase in deaths sufficient to have had an effect on life expectancy at younger ages (looking at life expectancy at birth and life expectancy at age 65)? For younger cohorts, life expectancy continues to increase slowly, as shown in figure 18, with increased mortality at age 85+ balanced by reductions in mortality in earlier years. However, as shown in figure 19, the rate of increase in life expectancy at birth has slowed substantially, and is now far slower than at any time in the last 35 years.Figure SEQ Figure \* ARABIC 18: Long-term trend in UK life expectancy at birthFigure SEQ Figure \* ARABIC 19: Long-term trend in annual rate of change in life expectancy at birthJust like life expectancy at birth, life expectancy at 65 is also continuing to rise (just), but the rate of increase has slowed substantially since 2011, and is now slower than at any point in the last 35 years, as shown in figures 20 and 21, below:Figure SEQ Figure \* ARABIC 20: Long-term trend in UK life expectancy at age 65Figure SEQ Figure \* ARABIC 21: Long-term trend in annual rate of change in life expectancy at age 65Exploring the increase: 7. Are there particular disease groups that have been responsible for the increase in deaths? As shown in figure 22, the principal change in recent years has been the increase in number of deaths coded to dementia and Alzheimer’s (up from 25,000 per annum in 2010 to 62,000 per annum in 2015). It is not clear how much of the increase in deaths coded to dementia and Alzheimer’s has been due to changes in coding practice (the government has incentivised higher diagnosis rates for dementia) and how much is due to actual change in cause of death.Apart from dementia and Alzheimer’s, trends in cause of death by other disease types have been quite gentle: there has been a small reduction in deaths from circulatory diseases, a small increase in deaths from cancer, and a small increase in deaths from respiratory conditions. Figure SEQ Figure \* ARABIC 22: Annual number of deaths in England and Wales by cause of deathExploring the increase: 8. Has the change in trend in life expectancy been picked up by policy-makers, for instance in the Office of Budget Responsibility’s forecasts of fiscal sustainability?In the Office of Budget Responsibility’s “Fiscal Sustainability Report” of January 2017, which used data from the October 2015 update of the Office for National Statistics’ population projections, OBR makes the comment that “… the ONS projections assume higher mortality rates, which reduces spending by 0.1 per cent of GDP in 2065-66. The cumulative impact of lower spending would reduce public sector net debt by 3.0 per cent of GDP by the same date; … under our interpretation of the longevity link on the SPA, the downward revision to life expectancy due to higher mortality rates would push back the profile of SPA rises … “The ONS publication from October 2015 (which takes account of 2012 – 2014 data on higher deaths, but predates data on 2015 and 2016 increases in deaths), projects continued increases in life expectancy at birth, but projects a slower rate of increase than had been assumed in ONS’s 2012 population projection. Potential causes of the increase: 9. Was the increase in deaths caused by growth in UK’s population, and/or by ageing of the UK’s population (e.g., baby boomer generation becoming elderly)The increase that we have seen after 2011 is quite different to what was predicted by ONS: when ONS published its “Subnational population projections 2012”, which included projections of number of deaths per year, this indicated that the long-term trend of annual decline in number of deaths would reach a minimum in 2016, before the effect of growing population and of the ageing of baby-boomers would lead to a slow increase in number of deaths from 2017 onwards, as shown in Figure 23, below. Figure SEQ Figure \* ARABIC 23: ONS’s projection of long-term trend in future deaths in England (part of 2012-based population projections)This projection is clearly materially different from the reality of 2011 being the minimum, and of a sharp increase in number of deaths over the period 2012 – 2016. As such, it appears that the cause of the actual increase seen is caused predominantly by factors other than the effects that were predicted in 2012 of growing population and of the ageing of the baby-boomer generation.Potential causes of the increase: 10. Is the increase an effect that has been seen across multiple comparator countries, or is it unique to the UK?The figures below show numbers of births and deaths over time in each of Spain, Germany and France. Note that the Eurostat data only goes up to 2014 (whereas our data for the UK goes up to 2016 for some metrics). Each of the figures has its own characteristics (it would be possible to hypothesise the impact of recession on Spain’s births from 2008 onwards), but none has a trend on deaths that – up to 2014 - appears similar to that seen in the UK.Figure SEQ Figure \* ARABIC 24: Long-term trend in Spanish births and deaths (Eurostat data)Figure SEQ Figure \* ARABIC 25: Long-term trend in German births and deaths (Eurostat data)Figure SEQ Figure \* ARABIC 26: Long-term trend in French deaths vs French birthsPotential causes of the increase: 11. Was the increase caused by climate effects, for instance by cold winters?Deaths in the UK are seasonal: there are more deaths in winter than summer, and some trend for very cold weather to increase death rates. Every year ONS publishes data on winter mortality, examining how large the excess of winter deaths relative to summer deaths is. Figure 27 (below) shows ONS’s latest chart on this matter, which shows that while winter 2012/13 was colder than other winters in recent times, other winters since 2011 have not been cold in the UK.The figure also shows that excess winter mortality hasn’t been particularly high since 2011 – as shown in figures 5 and 6, the growth in mortality has been a year round phenomenon, rather than a phenomenon which has hit winter disproportionately. We have found no evidence that the increases in deaths since 2011 have been related to particular weather events. Figure SEQ Figure \* ARABIC 27: ONS analysis of excess winter mortality in the UK related to mean winter temperature Potential causes of the increase: 12. Was the increase caused by a rise in “lifestyle diseases”, for instance obesity and/or diabetes, amongst the very elderly?Public Health England publishes data on obesity rates: for instance figure 28, below (source: Public Health England), shows the progressive increase in adult obesity in England over the last 20 years:Figure SEQ Figure \* ARABIC 28: Trend in adult obesity in EnglandPublic Health England’s data shows that this phenomenon applies across age groups (including the elderly), not just amongst younger age groups, as shown in figures 29 and 30:Figure SEQ Figure \* ARABIC 29: Prevalence of obesity and overweight, men in EnglandFigure SEQ Figure \* ARABIC 30: Prevalence of obesity and overweight, women in EnglandCombining the data in Figure 28 with that from Figures 29 and 30 suggests that obesity rates will have risen for each of the age groups, at least over a 10-year timeframe. Figure 31 shows that the long-term increase in obesity has driven a long-term increase in health risk in the UK population between 1993 and 2010. However, since 2010, the data does not indicate that obesity-associated health risk has grown further.Figure SEQ Figure \* ARABIC 31: Long-term trend between obesity and health riskTaking diabetes as one lifestyle-related disease specifically, the number of people diagnosed with diabetes continues to show year-on-year increases, rising from 1.4 million in the UK in 1996 to 3.5 million in 2014 (with an estimated further 1.1 million currently with undiagnosed diabetes). Potential causes of the increase: 13. Was the increase caused by a change in environmental factors, potentially associated with diesel and/or air pollution, since 2011?The Royal Colleges of Physicians and of Paediatrics and Child Health have estimated that air pollution is contributing to approximately 40,000 early deaths per annum in the UK. Public Health England has estimated that, in parts of the UK air pollution is a significant cause of 8% of deaths, as shown in the Figure 32 below:Figure SEQ Figure \* ARABIC 32: Estimated percentage of UK deaths attributable to air pollution, geographic spreadThis data shows air pollution contributing to a significant number of deaths. Data also shows that diesel (which contributes more to particulate pollution) progressively gained market share in the UK, reaching over 50% market share in 2014, before dropping back since then.However, this data does not yet show a clear trend over time in a way that is sufficient to link changes in air pollution to the increase in death rates seen since 2011. Potential causes of the increase: 14. Was the increase caused by a real terms reduction in social care expenditure since 2011?The Kings Fund has produced analysis that showed a 17% real-terms fall in spending on social care for older people between 2009/10 and 2013/14, as shown in Figure 33 below.Figure SEQ Figure \* ARABIC 33: Kings Fund analysis, spending on social care for older people (note: vertical axis units should be ? bn)Figure 34, which uses data from the National Social Care Intelligence Service, shows that this reduction in spending has led to a reduction in the number of people receiving Home Care from social services:Figure SEQ Figure \* ARABIC 34: NASCIS analysis of number of people receiving Home Care from social servicesAgeUK, in its “Care in Crisis 2014” report, shows that local authorities have tightened their eligibility criteria for funded care, as shown in Figure 35. As a result, by 2013/14, 87.5 per cent of authorities were only providing funded care for those with ‘substantial’ or ‘critical’ needs.Figure SEQ Figure \* ARABIC 35: Number of local authorities setting social care eligibility thresholds at each levelThis data does not prove that it was the reduction in provision of funded social care to older people was the cause of the increase in death rate seen since 2011, nor provide a quantification of how much reduction in social care provision contributed to the increase. However, the link is credible and worthy of further investigation.Potential causes of the increase: 15. Was the increase caused by declining NHS performance (e.g., A&E waiting times; delayed transfers of care; referral to treatment times) on a number of key measures since 2011?The Kings Fund, in its Quarterly Monitoring Report in November 2016, produced Figure 36 which shows the trend of increasing waiting times in A&E, since a stable low of waiting times that persisted between 2005/06 and 2009/10:Figure SEQ Figure \* ARABIC 36: Percentage of patients in Type 1 A&Es in England not meeting the standard of less than 4 hours from arrival to admission, transfer or dischargeIn the same Quarterly Monitoring Report, as shown in Figure 37 below, the Kings Fund demonstrated how the percentage of patients whose “referral to treatment time” waiting times for planned care have been longer than the 18 weeks standard, has been increasing:Figure SEQ Figure \* ARABIC 37: Kings Fund analysis of 18 week referral to treatment time performanceThe same report also shows how delayed transfers of care (i.e., patients whose discharge from hospital was delayed because of lack of availability of downstream health or social care services) have increased, as shown in Figure 38:Figure SEQ Figure \* ARABIC 38: Kings Fund analysis of delayed transfers of care, number on last day of each monthThere is evidence that links delays in treatment to higher mortality: this exists as case studies of individual deaths (for instance the two patients who died at Worcester Royal Infirmary following trolley waits over New Year 2017) and as more systematic academic research (for instance, it is known that delays within A&E/ ED cause crowding within the department; the review by Bernstein et al. (2009) found that “A growing body of data suggests that ED crowding is associated… with objective clinical endpoints, such as mortality”.)The NHS’s performance problems are certainly causing some additional deaths: however, the publicly available evidence that we have gathered so far does not yet provide a quantified link between the NHS’s performance problems and the absolute increase in mortality that has been seen since 2011. As with social care, the link appears credible and worthy of further investigation.Potential causes of the increase: 16. Was the increase caused by declining mental health within the elderly population, for instance related to anxiety, depression and loneliness?Nationally, the prevalence of diagnosed and recorded depression has been increasing (for instance, NHS QOF records indicate that the prevalence in England – across all ages – increased from 7.3% to 8.3% between 2014/15 and 2015/16).Prevalence and recorded depression is higher than 8.3% amongst the elderly, as shown in figure 39 below:Figure SEQ Figure \* ARABIC 39: Prevalence of diagnosed depression with age in England in 2014, based on GP and hospital records (Source: The Age UK almanac of disease profiles in later life, Melzer D et al, 2015)According to AgeUK’s “Loneliness and Isolation Evidence Review”, the Growing Older project (see GO programme website (Sheffield University/ ESRC): growingolder.group.shef.ac.uk) had isolation and loneliness as one of its 25 themes, with findings including that:7 per cent of older people were often lonely and 31 per cent sometimes lonely11–17 per cent were socially isolated in 2001These rates had remained relatively stable in the previous 50 years.A significant pattern is that both loneliness and isolation appear to increase with age, and among those with long-term health problems.This data indicates that depression, isolation and loneliness are all problems amongst the elderly. The data as presented so far does not (yet) demonstrate that these are increasing problems, or that there is an increase that is making a major contribution to the increased mortality that has been seen.Potential causes of the increase: 17. Was the increase caused by increasing inequality, and/or by increasing poverty within more deprived communities?The increase in mortality is concentrated in the 65+ cohort, and within this particularly the 85+ cohort, so one possible driver could be increasing poverty among pensioner households. Data on both absolute poverty (figure 40) and relative poverty (figure 41) show poverty rates flat or slightly declining for these groups. This suggests that rising poverty is not – on its own - a direct explanation for rising mortality.A second possibility is that declining poverty has driven falling mortality and increasing life expectancy for much of the last twenty years but that as the decline in poverty has flattened out, this effect has now been removed. As there are multiple long-term drivers of mortality this could create a situation in which drivers increasing mortality predominate, leading to the trend seen since 2011. This is a hypothesis at this stage and would need to be tested further.Figure SEQ Figure \* ARABIC 40: Absolute poverty rates for pensioners, children, working-age non-parents and all households (Source: Living Standards, Poverty and Inequality in the UK: 2016, IFS, 2016Figure SEQ Figure \* ARABIC 41: Relative poverty rates for pensioners, children, working-age non-parents and all households (Source: Living Standards, Poverty and Inequality in the UK: 2016, IFS, 2016) ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download