Rodney P Jones

• BSc Hon PhD
• Statistical Advisor at Healthcare Analysis & Forecasting

The Anthropocene, shaped by human forces, upends the concept of a stable, natural mortality baseline. We analyzed European monthly mortality data using the STL model to isolate annual cycles, then applied sparse PCA to identify key dimensions. For countries with data since 1960, Levene's test confirmed a convergence of mortality cycles toward a Eur...

Pediatric/neonatal bed demand are shown to depend on the trend in the number of births. Periods of higher births are demonstrated to generate a capacity shock, where length of stay (LOS) is reduced, sometimes by premature discharge, to squeeze arriving patients into existing bed capacity. Similar capacity shocks occur during the autumn/winter/sprin...

Optimum bed occupancy in hospitals is a disputed area. Some claim that an average occupancy of 85% applies. This short methodology study explains how to calculate the optimum occupancy in maternity units based on their size.

This study investigates the process of planning for future inpatient resources (beds, staff and costs) for maternity (pregnancy and childbirth) services. The process of planning is approached from a patient-centered philosophy; hence, how do we discharge a suitably rested healthy mother who is fully capable of caring for the newborn baby back into...

Rodney P Jones presents his perspective of hospital admission forecasting, explaining why deaths may be a more significant indicator than widely used forecasting methods allow.

This short article demonstrates why the English resource allocation formula which distributes funding to Integrated Health Boards is subject to a serious flaw. This flaw arises due to the fact that the resources allocation formula ignores the reality of the nearness to death (NTD) effect on healthcare costs. Distortions in supposed 'fair shares' fu...

This short article discusses how to correctly fund the English National Health Service (NHS) using the principles of nearness to death (NTD), sometimes called counting backward from death. NTD shows why the trend in the absolute number of deaths is so important to understand the cost pressures relating to healthcare.

Three models/methods are given to understand the extreme international variation in available and occupied hospital bed numbers. These models/methods all rely on readily available data. In the first, occupied beds (rather than available beds) are used to measure the expressed demand for hospital beds. The expressed occupied bed demand for three cou...

Overview: This study presents a method to assess the expressed bed demand (as occupied beds) in different countries. An initial comparison is made between Australia, England and the USA. Age standardized miortality (ASMR) is then used as a way of adjusting bed demand for population 'deprivation'. Finally, a method based on lines of constant turn-aw...

All vaccines exhibit specific and nonspecific effects. Specific effects are shown by the efficacy against the target pathogen, while nonspecific effects can be detected by the change in all-cause mortality. The real-world non-COVID-19 all-cause mortality (NCACM) data for the population of England between January 2021 and May 2023 was assessed. All...

Based upon 30-years of research by the author, a new approach to hospital bed planning and international benchmarking is proposed. The number of hospital beds per 1000 people is commonly used to compare international bed numbers. This method is flawed because it does not consider population age structure or the effect of nearness-to-death on hospit...

Based upon 30-years of research by the author a new approach to hospital bed planning and international benchmarking is proposed. The number of hospital beds per 1000 population is commonly used to compare international bed numbers. This method is flawed because it does not consider population age structure or the effect of nearness-to-death on hos...

Since 2020, COVID-19 has caused serious mortality around the world. Given the ambiguity in establishing COVID-19 as the direct cause of death, we first investigate the effects of age and sex on all-cause mortality during 2020 and 2021 in England and Wales. Since infectious agents have their own unique age profile for death, we use a 9-year time ser...

A recent study has suggested that the age profiles for deaths due to COVID-19 variants differs between variants and shows male/female specificity. This study implies that age/sex dependency is common among human pathogens. The often-reported higher susceptibility to infections among the young and elderly is true in general but does not apply to ind...

. The number of hospital beds per 1000 population is commonly used to compare international bed numbers. This method is flawed because it does not consider population age structure or the effect of nearness-to-death on hospital utilization. Deaths are also serving as a proxy for wider bed demand arising from undetected outbreaks of 3 000 species of...

This article discusses how unrecognised outbreaks of the 3,000 known species of human pathogens may be influencing the demand for healthcare. Such outbreaks probably act at local level. The blunt tools used in healthcare policy do not reflect such local pressures.

This short article investigates if certain vaccines may have unintended nonspecific effects against all-cause winter mortality, and hence upon the winter pressures experienced by the health services.

All vaccines exhibit both specific and non-specific effects. The specific effects are measured by the efficacy against the target pathogen, while the non-specific effects can be detected by the change in all-cause mortality. All-cause mortality data (gender, age band, vaccination history, month of death) between January 2021 and May 2022 was compil...

This study investigates the effects of age and sex on reported COVID-19 deaths and upon all-cause mortality during 2020, 2021 and 2022 to reveal the detail of the age-specificity of SARS-CoV-2 variants. The widely reported higher deaths in males was confirmed for most ages above 43 years in both 2020 and 2021. During 2020 (+ COVID but no vaccinatio...

A new model to compare international hospital bed numbers is presented. The key feature of the model is that it incorporates the absolute number of deaths (all-cause mortality). This is important for three reasons: 1. Around 55% of lifetime use of a hospital bed occurs in the last year of life, irrespective of the age at death 2. Any agent capable...

Pathogen interference is the ability of one pathogen to alter the course and clinical outcomes of infection by another. With up to 3000 species of human pathogens the potential combinations are vast. These combinations operate within further immune complexity induced by infection with multiple persistent pathogens, and by the role which the human m...

Comparing international or regional hospital bed numbers is not an easy matter, and a pragmatic method has been proposed that plots the number of beds per 1000 deaths versus the log of deaths per 1000 population. This method relies on the fact that 55% of a person’s lifetime hospital bed utilization occurs in the last year of life—irrespective of t...

This document provides a list of over 200 published studies in the area of health service research covering the forecasting of demand, financial risk in healthcare purchasing, limitations of the English HRG tariff, and calculating optimum hospital bed numbers and occupancy.

Unexpected outcomes are usually associated with interventions in complex systems. Excess winter mortality (EWM) is a measure of the net effect of all competing forces operating each winter, including influenza(s) and non-influenza pathogens. In this study over 2400 data points from 97 countries are used to look at the net effect of influenza vaccin...

I have reached the uncomfortable conclusion that neither the DHSC nor NHS E&I want the NHS to understand how to conduct correct capacity planning. Years of financial pressures has meant that CCGs simply fudge the numbers such that demand supposedly equals break even. This is embedded behaviour which I have consistently observed over the past 30 yea...

Trends in excess winter mortality (EWM) were investigated from the winter of 1900/01 to 2019/20. During the 1918–1919 Spanish flu epidemic a maximum EWM of 100% was observed in both Denmark and the USA, and 131% in Sweden. During the Spanish flu epidemic in the USA 70% of excess winter deaths were coded to influenza. EWM steadily declined from the...

Background: The current method for assessing critical care (CCU) bed numbers between countries is unreliable. Methods: A pragmatic method is presented using a logarithmic relationship between CCU beds per 1000 deaths and deaths per 1000 population, both of which are readily available. The method relies on the importance of the nearness to death...

Trends in excess winter mortality (EWM) were investigated from the winter of 1900/01 to 2019/20. During the 1918-1919 Spanish flu epidemic a maximum EWM of 100% was observed in both Denmark and the USA. During the Spanish flu epidemic in the USA 70% of excess winter deaths were coded to influenza. EWM steadily declined from the Spanish flu peak to...

German states most affected by COVID-19 • Germany was not greatly affected during the first COVID winter of 2019/20, however, there was significant excess winter mortality (EWM) during 2020/21. • Worst affected was Sachsen with 62% EWM. Sachsen shares a border with Poland and Czechia. • Next affected was Thüringen with 41% EWM. Thüringen is adjace...

Trends in excess winter mortality (EWM) were investigated from the winter of 1900/01 to 2019/20. During the 1918-1919 Spanish flu epidemic a maximum EWM of 100% was observed in both Denmark and the USA. During the Spanish flu epidemic in the USA 70% of excess winter deaths were coded to influenza. EWM steadily declined from the Spanish flu peak to...

Safest places to live during COVID • Safest places to live in the UK during COVID-19 have been Moray and Highland in Scotland, Torridge, Mid and East Devon in Devon, and Mendip in Somerset. • Worst CCG to live in was NW London, worst local authority – Brent. Ongoing trends in deaths in the UK • COVID-19 admissions per COVID deaths reaches a minimum...

Year of birth and increased deaths in 2020 for England and Wales • The year 2020 saw COVID-19 waves 1 and 2. • Wave 3 due to the Kent variant largely occurred in 2021. • Year of birth is used to capture potential birth cohort effects. • Year of birth effects are obscured using the 5-year age bands used in most analysis regarding the effects of age...

The first article in this two-part series demonstrated that the post-World War II ‘baby boom’ and increasing life expectancy implies that there will be a considerable increase in deaths in the UK over the next 40 years. As the highest proportion of lifetime hospital bed occupancy occurs in the last 1 year of life, this has significant implications...

Situation in the UK • Proportion of deaths due to COVID-19 continues to steadily rise, especially in Northern Ireland and London. • The slope of the rise in Northern Ireland is very similar to that at the start of the second wave! • Figure 1, Part 1 has been updated to week ending 27th August. All regions are showing an increase in the rolling 52-w...

Doctors are often asked to make input into bed calculations but are often not provided with the necessary background to the potential flaws in such calculations. A simple method is presented which allows both inter- and intra-national comparison of bed numbers which are sensitive to both population age structure and the role of nearness-to-death in...

The volume of COVID-19 testing in the world's countries shows an exponential decline that is correlated with GDP per capita and ranges from relative testing levels starting at a low of 3,300 cumulative tests per million persons (during the first 18 months of the pandemic) in the Democratic Republic of Congo (DRC) to a high of 13.7 million cumulativ...

The World War II baby boom, coupled with increasing life expectancy, will lead to increasing numbers of deaths for the next 40 years. The last year of life represents a large proportion (55%) of lifetime hospital bed occupancy. This is called the nearness to death effect. However, the nearness to death effect has not been factored into NHS capacity...

A method based on the underlying trend in deaths is used to verify excess deaths in the UK in both first and second waves • Genuine excess deaths were generally higher in the first wave. • The South East and Northern Ireland were the only exceptions. • London was highest in both waves. • Reported COVID-19 deaths were underestimated in the first wav...

International situation • Hot on the heels of Peru the Indian state of Bihar has dramatically revised COVID-19 deaths upward after a court challenge forced a review, Covid: India posts global record deaths after state revises data | India | The Guardian • The suspicion is that even the forced revision is a gross underestimate! o Even after revision...

International situation • “Reported” Covid-19 deaths have already passed 3.8 million. • The top 50 countries for increasing deaths in the past 7 and 14 days are listed. • Highest increase in Bahrain. • UK now ranks 122nd. The human microbiome and implications to COVID-19 • The immune system is vastly complex. • New types of immune cells are still b...

International situation • “Reported” Covid-19 deaths have already passed 3.7 million. • Peru revised its Covid death toll (on Monday) from 69,340 up to 185,000 (Yikes!) giving Peru a jump from 38% to 102% Covid excess deaths compared to deaths in 2019. • Expect to see more similar revisions in the coming months • Other badly hit countries now close...

India is a collection of federal states with each state responsible for health care, COVID-19 testing, and reporting of COVID deaths. Some states, which are mainly urban, are far more affluent than others and this greatly influences expenditure on COVID testing. COVID testing will also be mainly concentrated in easy to reach urban areas. The death...

International situation • “Reported” Covid-19 deaths are about to pass 3.5 million. • Indian states largely in exponential growth. o Highest in Lakshadweep at 500% growth in last 24 days – but small numbers. o Total for India is driven by Maharashtra (30% of total reported deaths in India), and only 30% growth in last 24 days. o Maharashtra has the...

Under-reporting of COVID-19 infections and deaths is as endemic as the virus itself. Undercounting is consistently associated with a large proportion of the population living in slums, associated low levels of health care resources, and a poorly functioning death reporting process. These conditions apply to most of Africa and large parts of Asia. A...

Over-counting of Covid-19 mortality in the UK • Reported Covid-19 deaths have to be compared to ‘excess’ deaths due to all-cause mortality. • In the early days of the pandemic testing was limiting and so reported Covid deaths were less than excess mortality • As testing increased the non-specific definition of a Covid death then led to over-countin...

India has 1.9 million hospital beds yet is in the news for people unable to access hospital care for severe COVID-19 infection. “Reported” COVID-19 deaths for India are insufficient to explain this conundrum. This article is part of a wider series investigating gross under counting of COVID-19 mortality around the world, and how nearness-to-death d...

Hospital bed numbers in India • India has an estimated 1.9 million hospital beds ranging from state-of-the-art private hospitals to less-well-equipped rural hospitals. • Hospital beds are provided by multiple players including the military, the railways, a state insurance scheme for civil servants, state governments and the private sector. • Averag...

As of May 1st, 2021 the “reported” COVID-19 death toll stands at >3.2 million. The 3.2 million total had been exceeded by 8pm (GMT) of that day. Is the real death toll far higher than this? Low levels of testing for COVID-19 are common in around half of world countries. All these countries share common features, namely, high proportion of the popul...

Covid-19 deaths in India • India is a federal collection of 40 states and territories. • Each state/territory is far larger than many countries. • Bihar, the largest state by population, has an estimated 124.8 million people and only 2,560 “reported” Covid-19 deaths. One assumes a vast underestimate. • Lakshadweep, Daman and Diu, Dadra and Nagar Ha...

Part 34 of this series attempts to quantify the extent of undercounting of Covid-19 deaths around the world. A method based on reported Covid-19 mortality per 1,000 deaths (before Covid-19) plus extent of Covid-19 testing per million population indicates that half of world countries could be undercounting by up to a factor of 10-times. Undercountin...

Reporting of any sort of death in most of Africa is a lottery. • Only South Africa and Egypt have compulsory death notification. • The rest of Africa is questionable and is mostly paper records only. • 45% excess mortality in South Africa during the first peak was higher than the UK which at that time was just 14%! • Global Covid-19 deaths will be...

To investigate the dynamic issues behind intra- and international variation in EWM (Excess Winter Mortality) using a rolling monthly EWM calculation. This is used to reveal seasonal changes in the EWM calculation and is especially relevant nearer to the equator where EWM does not reach a peak at the same time each year. In addition to latitude coun...

Death acts as a dual proxy for nearness to death (NTD) and related wider morbidity. Hospital bed use in the last year of life accounts for somewhere around 25 bed days of hospital resource consumption. The knock-on morbidity due to the agents promoting death also accounts for up to another 25 bed days of hospital resource consumption. This is appro...

Public health agencies have promoted a single pathogen view of infection which simplifies the message about vaccination. However, research over the past 20 years shows a far more complex set of interactions between multiple pathogens, the nose/throat microbiota, the gut microbiota, immune function, and ageing. Influenza vaccination is effective at...

This series introduces epidemiological and medical concepts into understanding why health care costs and capacity pressures can be so volatile. COVID-19 is used as an example for many of the principles. The role of the nearness to death (NTD) effect in healthcare demand is discussed in relation to acute hospital bed utilization in U.S. states. The...

To investigate the dynamic issues behind international variation in EWM. A rolling/moving EWM calculation is used to reveal seasonal changes in the EWM calculation and is especially relevant nearer to the equator. In addition to latitude country specific factors determine EWM. Females generally show higher EWM mainly due to respiratory conditions....

More than 81,000 excess deaths (all-cause mortality) occurred in the UK during 2020 • Highest excess occurred in London (+21.9%) and lowest excess occurred in the South West (+7.4%). Northern Ireland consistently rose to roughly the same as Wales and the South East, and eclipsed Scotland (+10.3%). • Ranking of the regions is roughly according to we...

Around 80,000 excess deaths are likely for the UK during 2020 • As predicted in Part XXIV the trend for >80,000 (possibly >82,000) excess deaths by the New Year remains unchanged • After a full-year of Covid-19 (end of March 2021) up to 100,000 excess deaths are possible • Only the North West shows preliminary signs of the effects of lock down on a...

The weighted population density for 349 English local government areas, and the raw population density for a further 58 areas from Northern Ireland and Scotland, were used to demonstrate the role of high population density in the volatility associated with year-to-year total deaths and of excess winter mortality (EWM). Volatility in EWM was measure...

The financial and capacity pressures experienced by hospitals and social care organisations are far higher than has been realized. End-of-life is a time of high utilization of acute and social care. The absolute number of deaths (and its year-to-year variation) then acts to drive the marginal pressures in both capacity and costs for these organisat...

Covid-19 shows slow/fast spread behaviour, i.e., spatiotemporal granularity • The maximum potential for Covid-19 deaths appears to be a 40% to 50% increase in all-cause mortality • Thankfully, this requires extremely high population density, as in parts of New York and London, to achieve • Many local authorities (usually with lower-than-average wei...

Around >80,000 excess deaths are likely for the UK during 2020 • This is around 11,000 higher than the official Covid-19 deaths • This gap is partly due to low testing capacity in the early months of the epidemic • Some may be due to the unintended effects of lock-down • My suspicion is that the subclinical effects of Covid-19 are more important th...

In the UK, government austerity has been suggested as a major reason for the stalling of life expectancy improvements and age-standardised mortality rates. However, these trends have also been observed in many other countries. Influenza has been suggested as a potential confounding factor, as this condition contributes significantly to excess winte...

A new model for hospital bed numbers which adjusts for end-of-life care and age structure is used to demonstrate that England has 20% fewer occupied beds that the other countries in the UK. It also shows that occupied beds in English Clinical Commissioning Groups (CCGs) lie parallel to a line of equivalence with New Zealand and Singapore. This is d...

It is widely recognised that the Covid-19 death rate increases exponentially with age. However, somewhat paradoxically, countries with the oldest populations tend to have the lowest excess Covid-19 deaths. This conundrum is explained by the simple fact that spread of Covid-19 is mainly due to young adults. Young adults will mainly have asymptomatic...

In the UK, government austerity has been suggested as a major reason for the stalling of life expectancy improvements and age-standardised mortality rates. However, these trends have also been observed in many other countries. Influenza has been suggested as a potential confounding factor, as this condition contributes significantly to excess winte...

In this analysis total deaths are used as a proxy for Covid-19 as it is often difficult to determine exactly whose death is directly due to Covid-19 rather than death from other causes along side Covid-19 infection. As the old nursery rhyme goes "Who killed cock robin?" As always, deaths are a lagging indicator, which for many infectious outbreaks...

Data from UK local and regional government areas at the end of August is used to demonstrate the relatively slow spread of Covid-19 throughout the UK. • This is entirely like that demonstrated in the USA in Parts XVI and XVII • Maximum deaths had been reached in 11% of local government areas by April, mainly in parts of London or other sporadic loc...

Data from the USA is used to illustrate the fact that certain communities have seen very few Covid-19 deaths and are probably highly susceptible to a 'second' wave'. The city of San Francisco is one example which in mid-September had only experienced 1.5% excess Covid-19 deaths relative to 2018 total deaths. Norfolk City in Virginia is another exam...

Data from the USA is used to illustrate the fact that there is no second wave, only a single wave gradually spreading to susceptible communities. The travelling wave took five months to reach all counties in the state of Alabama. Seven US states did not experience the maximum rate of Covid-19 deaths until September - the travelling wave effect. Ten...

Covid-19 has exposed huge deficiencies in the process of identification and coding of cause of death. • A preliminary analysis of deficiencies in the date of registration was given in Part VI and deaths in care homes due to Covid-19 were seemingly massively underestimated • See Part VIII for an estimate of undercounting in the early days of the epi...

Background/Aims While many methods exist for calculating bed numbers, the current method for making international comparisons is inadequate, as it makes no adjustment for population age or the nearness to death effect. This study investigated the number of acute medical beds in various countries using a new method to compare bed numbers within and...

Covid-19 deaths can serve as an international proxy for hospital capacity pressures using the ratio of Covid-19 deaths per total hospital beds. Detailed comparison of potential capacity pressures between the states in the USA has been published, see https://www.mdpi.com/1660-4601/17/14/5210/pdf • State weighted population density was correlated wit...

Mid super output areas (MSOA) in England and Wales contain around 7,900 persons (range 2,200-24,900). There is a huge range in the small-area proportion of deaths due to Covid-19 for the period March to June 2020. The proportion of total deaths due to Covid-19 is highly dependent on population density. Some15% of deaths were due to Covid-19 for lea...

Mid super output areas (MSOA) in England and Wales contain around 7,900 persons (range 2,200-24,900). There is a huge range in the small-area proportion of deaths due to Covid-19 for the period March to June 2020. Mid super output areas (MSOA) in England and Wales contain around 7,900 persons (range 2,200 – 24,900). The median value was 21.4% Covi...

1) Background: To evaluate the level of hospital bed numbers in U.S. states relative to other countries using a new method for evaluating bed numbers, and to determine if this is sufficient for universal health care during a major Covid-19 epidemic in all states (2) Methods: Hospital bed numbers in each state were compared using a new international...

During the Covid-19 epidemic some areas show apparent 'negative' excess mortality which is not due to statistical randomness. This was confirmed using two different methods. These arise from the timing of Shift-Up/Down events in total deaths, which can be traced back to the 1950's (and probably before). A significant role for shift-up/down in death...

A curious series of Shift-Up/Down events in total deaths can be traced back to the 1950's (and probably before). In Shift-Up deaths are elevated in all months for a minimum of 12 months. The % increase in deaths at local authority level shows the same range as that for Covid-19, although Covid-19 kills people over a short period of time. Shift-Up e...

This is Part IX of an extended series. Covid-19 deaths in various countries are compared with hospital bed numbers as a means of determining the capacity shock potentially experienced in different countries. The UK ranks high in most measures.

Due to under-reporting of with-Covid-19 deaths it is 'excess' all-cause mortality which gives the true impact of the epidemic. However, deaths are always a lagged indicator of an infectious outbreak. The rolling 52-week total of deaths peaked at the end of May. See Fig. 1, Part I at the end of this document. Projected deaths beyond May seem to indi...

Due to under-reporting of with-Covid-19 deaths it is 'excess' all-cause mortality which gives the true impact of the epidemic. Deaths are always a lagged indicator of an infectious outbreak. As at the end of April (it is now June) the Brent Borough Council in London had experienced a huge 30% increase in excess mortality. The worst affected areas w...

Due to under-reporting of with-Covid-19 deaths it is 'excess' all-cause mortality which gives the true impact of the epidemic. There was a serious problem early in the epidemic due to the lag between date of death and date of registration., i.e. in England and Wales at week 11 only 5 with-Covid-19 registrations had been made but 47 with-Covid-19 de...

Due to under-reporting of with-Covid-19 deaths it is ‘excess’ all-cause mortality which gives the true impact of the epidemic. Excess deaths in England and Wales up to week 18 were: 21,750 in care homes, 18,600 in hospital and 11,250 at home, hospice and elsewhere. Curiously, the 24,685 with Covid-19 deaths recorded in hospital are 6,090 higher tha...

Due to under-reporting of with-Covid-19 deaths it is 'excess' all-cause mortality which gives the true impact of the epidemic. Under-reporting of with Covid-19 deaths was occurring in the early weeks of the outbreak, i.e. the outbreak was happening faster than first appeared. Data from Scotland indicates that actual deaths (by date of death) are ar...

Up to the 24th April in the UK for every 1,000 official Covid-19 death (deaths in all places) there were 550 extra deaths which could not be accounted for. Deaths in London had already reached the UK equivalent of 100,000 excess deaths. Deaths are compared to a large increase in deaths which occurred in 2014/2015. It is highly likely that the total...

This analysis looks at the trajectories of excess deaths across the UK during the Covid-19 epidemic. It attempts to align each region with an estimated start date for the outbreak in that region.

This is part of a series of preprints investigating the total number of 'excess' deaths which have occurred in the UK during the time of the Covid-19 epidemic. Total deaths are roughly 2-times the number of in-hospital deaths. Total deaths will include with Covid-19 deaths in nursing homes and at home, but will also include any unintended consequen...

This is an initial estimate of the total extra deaths due to Covid-19 in the UK. This will be refreshed weekly.