Kidney damage 'killing thousands,' study claims

Wednesday April 23 2014

“Failures in basic hospital care are resulting in more than 1,000 deaths a month from … acute kidney injury,” The Independent reports. A study commissioned by the NHS estimates that up to 40,000 people may be dying from this preventable condition.

The study aimed to discover the prevalence of acute kidney injury (AKI – previously called acute kidney failure) among adult inpatients in NHS hospitals.

AKI is characterised by a rapid decline in kidney function, which can have many underlying causes. The condition can have a high risk of multiple organ failure and death.

The researchers used data from the Hospital Episode Statistics (HES), which covers all NHS hospital admissions in England. They compared this with more detailed information on AKI obtained from three Kent hospitals to see whether the overall HES data was giving a reliable indication of the true prevalence of the condition in NHS hospitals.

The results suggest that the prevalence of AKI among hospital inpatients could be much higher than previously thought.

Overall, it was estimated that around 14% of hospital inpatients could have AKI. The mortality associated with this is also high – accounting for an estimated 40,000 inpatient deaths over any given year.

Previous research has suggested that around 20-30% of AKI cases could be prevented, and the study highlights the importance of recognising people who could be at risk of developing the condition.

The health watchdog NICE published guidelines on AKI in 2013.

Where did the story come from?

The study was carried out by researchers from Insight Health Economics (London); East Kent Hospitals University NHS Foundation Trust (Canterbury); NHS Improving Quality (Newcastle upon Tyne); and Salford Royal NHS Foundation Trust (Salford). Funding was provided by NHS Kidney Care.

The study was published in the peer-reviewed medical journal Nephrology Dialysis Transplantation and has been made available on an open access basis, meaning it is free to read online.

All the media headlines have focused on the angle that thousands of people are dying of thirst due to alleged poor care. This has been taken from the “avoidable” aspect of this acute kidney injury – where previous research (specifically the previous National Confidential Enquiry into Patient Outcome and Death) has shown that up to a third of cases could be prevented.

However, the research itself only looks at the prevalence, costs and outcomes of AKI.

It does not focus on identifying possible reasons for the high number of cases or ways they could be avoided.

Based on the evidence made available in the study, claims that 40,000 people are “dying of thirst” are unsupported.

What kind of research was this?

This was a modelling study, which had a number of related goals:

  • Examining the prevalence of AKI across the NHS.
  • Estimating the impact that AKI has upon mortality, length of hospital study, quality of life and healthcare costs.

Acute kidney injury (AKI), previously termed acute kidney failure, is the term used to describe when there is sudden damage to the kidneys. There is no widely accepted standard definition of AKI and there may be a number of difference causes.

Criteria tend to be based upon:

  • A sudden rise of blood creatinine levels above a certain threshold level (creatinine is a breakdown product produced by the muscles, and is a good indicator of kidney function).
  • A decrease in urine output below a certain threshold level.

It is a serious illness that has a high mortality risk, though the specific mortality risk will be highly variable, depending on the individual (such as whether there are complications, or the person has existing kidney damage or other medical problems).

Importantly, as previous research has highlighted, there are concerns that many cases of AKI could be prevented, which would cause considerable reductions in illness, deaths and healthcare costs. The 2009 National Confidential Enquiry into Patient Outcome and Death (NCEPOD) found that around a third of AKI cases occurring while in hospital were avoidable. Furthermore, only half of patients with AKI received an overall standard of care that was considered “good”.

This modelling study using fairly reliable data on NHS hospital admissions and is valuable research for estimating health outcomes of AKI and costs to the NHS.

What did the research involve?

This study used routinely collected national data for the NHS in England, in order to look at the prevalence of AKI in adults. They then estimated the impact of AKI on mortality, other health outcomes and cost to the NHS.

The researchers used Hospital Episode Statistics (HES), which is derived from records for each patient admitted to each NHS hospital. HES data includes the patient’s demographic details and medical information, including diagnoses, procedures, length of stay and in-hospital mortality.

They looked at recorded diagnoses of AKI (according to the International Classification of Diseases) between 2010 and 2011. 

However, the HES data does not include information on patients' AKI stage, kidney function prior to hospital admission or kidney function after being discharged from hospital.

As the researchers say, AKI is often poorly recorded in patient notes, so the national findings were compared with data collected by the three hospitals of the East Kent Hospitals University NHS Foundation Trust (EKHUFT).

This involved looking at laboratory records and identifying cases of AKI based on blood creatinine levels, using the Acute Kidney Injury Network (AKIN) classification system.

Comparing these two sources of information, they estimated the under-recording of AKI in patient notes.

They also used both data sets to estimate the possible distribution of AKI cases across the NHS according to stage, and to estimate the person’s previous and future kidney function.

They then used statistical models to estimate the impact of AKI on mortality, number of days in critical care and overall hospital stay.

What were the basic results?

AKI prevalence

The HES data indicated that AKI was recorded for 2.4% of hospital admissions during 2010/11 (142,705 out of 3,792,951 admissions). Prevalence ranged from 0.3% of patients aged 18 to 39, to 5.7% of people aged ≥80.

During the six-month period of EKHUFT data, laboratory research indicated that AKI was present in 15% of admissions, though the EKHUFT population is older than the overall HES population. 

When standardising for age, it was 14% of admissions. 

Over a third of patients (38%) in EKHUFT who had AKI during the study period had pre-existing chronic kidney disease. Three-quarters of those had AKI when they were admitted to hospital, suggesting that their condition was not due to poor hospital care.

AKI mortality

Using the HES data, just over a quarter (28%) of people with AKI recorded during their admission died before hospital discharge. The odds of in-hospital death were 10-fold greater in a person with AKI compared to those without. Mortality rates increased with age. 

From EKHUFT data, it was shown that 14% of people with AKI died before being discharged from hospital. In over half of all in-patient deaths during the six-month study period, the person had AKI recorded.

Analysis from HES data suggests that AKI was associated with around 15,000 excess deaths among inpatients in England in 2010/11.

However, extrapolating from EKHUFT data suggests that the number of excess inpatient deaths associated with AKI in England may be above 40,000.

Length of hospital stay

When using HES data, the average duration of hospital stay was 16.5 days for AKI-admissions, compared to just 5.1 days for admissions without AKI recorded. A person with AKI had a length of stay 2.6 times longer than someone without AKI; using the EKHUFT data, it was 1.6 times longer. From the EKHUFT critical care information, 60% of critical care bed days over the period were in people recorded to have AKI.

Long term outcomes and costs

Post-discharge information was not available from HES; using the EKHUFT data, 0.56% of people with AKI were receiving renal replacement therapy (such as dialysis) at 90 days, though more than half had pre-existing chronic kidney disease. 

Using the HES data, there were estimated to be almost 1,000,000 excess bed days due to AKI.

Based on the EKHUFT data, the number of excess bed days may be as high as 2.5 million, with over 160,000 of these spent in critical care beds. The total inpatient costs of AKI recorded in HES was estimated at £380 million.

When extrapolating from EKHUFT data, the cost could be as high as £1.02 billion – just over 1% of the NHS budget. To put that figure into context, that is enough to hire an additional 47,500 trainee nurses.

The lifetime costs of post-discharge care for people with AKI during admission was estimated at £179 million, with a loss of 1.4 quality of life years for each person with AKI who was admitted to hospital.

How did the researchers interpret the results?

The researchers conclude that the prevalence of AKI among people admitted to hospital may be considerably higher than previously thought, and up to 80% of cases may not be adequately captured by routine hospital data. AKI is associated with large numbers of in-hospital deaths and with high NHS costs.


This valuable study provides an estimate of the likely prevalence of AKI among inpatients in NHS hospitals. Comparison of HES data with laboratory data obtained from the three EKHUFT hospitals (where the AKIN classification system was used to define AKI cases), suggests that prevalence could be much higher than thought, and that there could be considerable under-recording of cases in the NHS.

The study also highlights the high mortality associated with AKI – accounting for an estimated 40,000 excess inpatient deaths. AKI was also associated with considerable loss to quality of life. Looking at the financial burden, this study estimated that AKI accounted for just over 1% of the NHS budget in 2010/11.

However, the study had its limitations. These figures are based on estimates only and centred on extrapolating data for HES based on data from the three EKHUFT hospitals. As noted, these hospitals have a different patient demographic from all NHS hospitals across England as a whole. There was also a lack of longer-term outcome data beyond 90 days after a patient was discharged from hospital.

Also, as the researchers say, this study only provides information on AKI recorded for adult hospital inpatients. There is no information on the number of cases that develop in the community.

The media has focused on the “preventable” aspect of AKI. Previous NCEPOD data has reported that up to a third of AKI cases could be predicated and prevented.

The researchers discuss how many of the failings identified by this report related to omissions in basic medical care. These include performing regular observations, checking the person’s fluid and mineral (electrolyte) balance, and a lack of adequate senior review. However, though the researchers mention fluid balance, at no point in this research paper do they say that “thousands are dying because of thirst”.

Notably, based on EKHUFT data, AKI was present at the point of admission in 75% of admissions where it was recorded, possibly noting a point for early recognition and management.

As the researchers say: “if 20% of AKI cases were prevented, the figures presented in this report suggest that the gross savings to the NHS could be in the region of £200 million a year, equivalent to 0.2% of the NHS budget in England”.

Analysis by Bazian
Edited by NHS Choices