Adverse events in British hospitals: preliminary retrospective record review
BMJ 2001; 322 doi: https://doi.org/10.1136/bmj.322.7285.517 (Published 03 March 2001) Cite this as: BMJ 2001;322:517
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
If, say, 10% of all admissions to medical facilities lead to an
adverse event,and say, 10% of these result in a fatality, then the chance
of death due to such an event is one in a hundred per person-admission (or
treatment). That is, a probability of 0.01 on average (often expressed
mathematically as 1E-02).
If a train commuter was exposed to the same risk as this due to
adverse events on the railways, then the result would be a fatality rate
of one per hundred passenger-journeys. Thus if the commuter made, say,
twenty journeys per month, his life expectancy would be about five months
(slightly longer than a Battle of Britain fighter pilot). This would
quickly put paid to rail travel.
In fact the chance of dying on a rail journey equates to about 1
fatality per 500,000 passenger-journeys or a probability of 0.000002 often
expressed as 2E-06. Travel by car doubles this risk per journey, and by
air the risk is about 6 times greater per journey than by rail (note that
aviation safety statistics are massaged to look better by using a
kilometre travelled basis to calculate the risk, comfirming most people's
gut feeling that air travel is inherently risky!).
Crudely stated, the chance of dying as a result of entering hospital
is therefore 5000 times greater than on entering a train (2500 times worse
than by car and about 1000 times worse than by aircraft per journey).
Given all the recent negative publicity about rail safety, this comparison
provides, to say the least, cause for reflection on the reasons for
adverse events in medical facilities.
What happens in a hospital to make it such a dangerous place compared
with a train (car, aircraft, etc)?
I am currently writing a book about fatal accidents resulting from
human error, including medical errors (or adverse events),but in
particular looking at the system faults which lead to errors (so-called
systemic or system induced errors).
Any suggested answers to the above question would be most welcome.
Barry Whittingham
Competing interests: No competing interests
The studies on adverse events associated with medical care only deal
with those invovling a hospital record. We have no data on events that
don't interface with a hospital admission. A substantial portion of these
may result in death, and not be associated with a hospital admission. The
hosptial deaths may only be the tip of the iceberg.
In the United States, and in most poor countries around the globe, we
need to expand the definition of an adverse event. This can include the
financial effect on a family of having to pay a medical bill. In my
experience working in Nepal, where private practice tends to be very cost-
intensive, and results in the ordering of many expensive tests of dubious
need. The patient and her/his family have no way of deciding whether the
expense is justified. Families will borrow money to pay for the care and
then limit spending on basic needs such as food, with resulting harm. The
same happens to families in the United States who incur huge medical
costs.
The United States spends 42% of the world's health budget (WHO 2000),
and ranks behind all other rich countries in what I call the Health
Olympics (ranking of countries by life expectancy). We should seriously
question what we buy with health care.
Competing interests: No competing interests
Dear Sir,
Re: Adverse events in British hospitals: preliminary retrospective
record review. BMJ, 322; 517-519
The authors of a recent paper commented that 'the epidemiology of
adverse events has not been studied in Britain.'1 Neither the paper or
the additional information on the BMJ Website provides detailed
information on what constitutes an adverse event. However, it would
appear from the example given, that hospital-acquired infections (HAIs)
are included in this category. Assuming this to be the case, it would
have been interesting to know whether all HAIs were classified as adverse
events caused by medical management, or whether some infections were
excluded because they were viewed as unfortunate consequences of the
disease process. If all HAIs were included in this category then there is
information on both the epidemiology of this particular adverse event,2
and the economic burden imposed.3
The paper reports that 46% of the adverse events identified were
judged preventable, and that preventable events cost the NHS around £1bn
per year in terms of additional bed days.1 It would be interesting to
know what proportion of these preventable events were HAIs and how this
judgement was made. Recent subjective estimates suggest that 15% of HAIs
could be prevented through improvements in infection control,4 earlier but
more objective data suggests that it might be twice this figure.5 If 15%
were prevented then, based on the results of recent estimates of the
economic burden of HAIs, the prevention of this type of adverse event
alone would result in the release of at least 546,000 bed days and
resources valued at £150 million. These estimates are limited to HAIs
occurring in adult non-day case patients admitted to selected specialties
of NHS Hospitals in England (approximately 70% of adult non-day case
admissions).3 The overall number of bed days and resources released from
the prevention of this type of adverse event is therefore likely to be
considerably higher.
1. Vincent C, Neale G, Woloshynowych M. Adverse events in British
hospitals: preliminary retrospective record review. BMJ 2001;322:517-519.
2. Emmerson AM, Entstone J, Griffin M, Kelsey MC, Smyth ETM. The
second national prevalence survey of infection in hospitals - Overview of
the results. J. Hosp Infect 1996(32):175-190.
3. Plowman R, Graves N, Griffin M, Roberts JA, Swan AV, Cookson B,
et al. The rate and cost of hospital acquired infections occurring in
patients admitted to selected specialties of a district general hospital
in England and the national burden imposed. J Hosp Infect 2001;47(3):198-
209.
4. Report of the Comptroller and Auditor General. The Management and
Control of Hospital Acquired Infection in Acute NHS Trusts in England.
London: The Stationary Office Limited, 2000.
5. Haley RW. Managing hospital infection control for cost-effectiveness:
a strategy for reducing infectious complications. Chicargo: American
Hospital Publishing, 1986.
Rosalind Plowman
Lecturer
Department of Public Health and Policy,
London School of Hygiene and Tropical Medicine,
Keppel Street,
London WCIE 7HT
Email: R.plowman@ lshtm.ac.uk
Jennifer A Roberts,
Reader in the Economics of Public Health
Department of Public Health and Policy,
London School of Hygiene and Tropical Medicine,
Keppel Street,
London WCIE 7HT
Nicholas Graves,
Lecturer,
Department of Public Health and Policy,
London School of Hygiene and Tropical Medicine,
Keppel Street,
London WCIE 7HT
Mark A S Griffin
Lecturer in Medical Statistics
Royal Free & University College Medical School,
Department of Primary Care & Population Sciences,
Level 2,
Archway Campus,
Holborn Union Building,
Highgate Hill,
London N19 3UA
Barry Cookson,
Director of the Laboratory of Hospital Infection,
Central Public Health Laboratory,
61, Colindale Avenue,
London NW9 5HT
Lynda Taylor
Head of Infection Control Unit,
Central Public Health Laboratory,
61, Colindale Avenue,
London NW9 5HT
Competing interests: No competing interests
Sir,
Vincent reports that 10.8% of patients experienced an adverse event
in hospital (1). There has been increasing recognition about the
importance of medical errors and what can be learnt from them (2). Rather
than undertaking yet further research to describe the scale of the
problem, efforts would be better invested in designing and evaluating
effective interventions to reduce it.
Errors in diagnostic radiology have been recognised and analysed for
many years (3,4). The Royal College of Radiologists has recommended errors
meetings for radiologists to discuss mistakes and learn from them. We
reviewed the results of two years of self-reporting of radiological errors
in Bradford to determine how many resulted in litigation or adverse
clinical outcomes.
Approximately 200,000 examinations were reported annually by 9
radiologists. During 1998 and 1999 all reported errors that were
identified from repeat examinations or by clinicians were reviewed by one
of the authors (SC). Minor errors were discarded. The remainder were
reviewed anonymously in errors meetings and the lessons discussed.
Complaints and litigation cases in the department were also reviewed for
the years 1998 to 2000.
35 major errors were reviewed by all radiologists. The majority were
reported by the individual who made the error. None resulted in complaint
or litigation. In addition none of the complaints or legal cases during
this period could be traced back to reported errors.
One of the major justifications for risk reporting and management
strategies is to reduce complaints and malpractice claims. However we
found no evidence that a self-reporting system achieved this. The number
of identified errors reported was small in comparison to the number of
examinations carried out. Experience of errors meetings from other
hospitals suggests that this number is typical. It is inevitable that
there were many other errors that were not identified or not reported.
The errors meetings did lead to the implementation of two changes in
practice, which were minor issues of protocol. They also provided a forum
for education and debate. However it is unclear whether they resulted in
better practice or reduced subsequent errors. Claims to the contrary
should be evidence-based. In the current NHS climate, it may be that the
wrong errors continue to be reported.
S. Chakraverty
Consultant Radiologist
Ninewells Hospital and Medical School,
Dundee DD1 9SY
sam.chakraverty@tuht.scot.nhs.uk
Dr. J. Wright
Associate Medical Director
Consultant in Epidemiology and Public Health,
Bradford Hospitals NHS Trust,
Duckworth Lane,
Bradford BD9 6RJ
john.wright@bradfordhospitals.nhs.uk
1. Vincent C, Neale G, Woloshynowych M. Adverse events in British
Hospitals: preliminary retrospective record review. British Medical
Journal (2001) 322:517-9.
2. Department of Health. Organisation with a memory. Report of an
expert group on learning from adverse events by the Chief Medical Officer.
London:HMSO 2000.
3. Smith MJ. Error and Variation in Diagnostic Radiology. Pub.
Charles C Thomas, Springfield, Illinois, USA , 1967. Library of Congress
Catalog Card No. 67-116118.
4. Robinson PJ. Radiology’s Achilles’ heel – error and variation in
the interpretation of the Rontgen image. British Journal of Radiology
(1997) 70:1085-98.
No competing interests.
Competing interests: No competing interests
In a recent survey, Vincent et al. [1] have estimated that
approximately 11% of hospital admissions in two British hospitals were
associated with an adverse event. The authors argue for a larger scale
study to document the prevalence of medical adverse events in the UK.
However, their estimate of in-hospital adverse events is well within the
statistical boundary of previous estimates obtained by much larger US [2]
and Australian [3] studies. Therefore, their argument for yet another
large study seems weak and perhaps irrational, particularly in light of a
recent UK study which has documented the problem of substandard care [4].
We believe that there is sufficient and necessary evidence to state that
medical error is common; and that there is a urgent need to develop
strategies to prevent or reduce the error, not to spend resources for
conducting another descriptive epidemiological study.
Among the adverse events, death, cardiac arrest and unplanned
admissions to intensive care unit (ICU) are probably the most serious
occurrences. Most of those events have their genesis in the general
wards. The events are not sudden nor unpredictable, because they are
usually preceded by signs of clinical instability. Recognising these
facts, a system called the Medical Emergency Team (MET) has been developed
and trialed since 1990 [5]. The system has three components: identifying
high-risk patients at an early stage; providing a rapid appropriate
response to the patients; and providing feed-back data on the
effectiveness of the MET system. Under the MET system, when a patient's
clinical condition is unstable (as judged by specific criteria), a call is
immediately placed to the MET for intervention. We postulate that the MET
system is an effective strategy to reduce in-hospital adverse events.
In a prospective study conducted in a 300-bed tertiary referral
teaching hospital in Melbourne, after implementing the MET system, the
incidence rate of cardiac arrest was reduced by 50% (unpublished data).
Moreover, in a recent six-month prospective study, the incidence rates of
cardiac arrest, deaths, and unplanned ICU admissions in one MET-based
hospital was lower than those in the two hospitals without the MET system;
however, after adjusting for case-mix factors, a statistically significant
difference was only observed in the rate of unplanned ICU admissions [6].
Nevertheless, further randomised controlled studies with multiple
hospitals are required to systematically evaluate the effectiveness of the
MET system. Such a study is underway in Australia and New Zealand, and we
suggest that a similar strategy and study be considered by the National
Health Service.
Tuan V. Nguyen1,2, Michael D. Buist3, Ken M. Hillman1,2
1Department of Anaesthetics, Intensive Care and Emergency Medicine,
University of New South Wales, Sydney;
2The Simpson Centre for Health Service Innovation, Liverpool Hospital,
Sydney;
3Intensive Care Unit, Dandenong Hospital, Victoria, Australia.
Correpondence author:
Dr. Tuan V. Nguyen,
The Simpson Centre for Health Service Innovation,
Liverpool Hospital,
Elizabeth Street, NSW 2170
Australia.
Email: t.v.nguyen@unsw.edu.au
References
1. Vincent C, Neale G, Woloshynowych. Adverse events in British
hospitals: preliminary retrospective record review. BMJ 2001; 322:517-9.
2. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG,
Newhouse JP, Weiler PC, Hiatt HH. Incidence of adverse events and
negligence in hospitalized patients. Results of the Harvard Medical
Practice Study I. N Engl J Med 1991; 324:370-6.
3. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Newby L, Hamilton JD.
The Quality in Australian Health Care Study. Med J Aust 1995;163:458-71.
4. McQuillan P, Pilkington S, Allan A, Taylor B, Short A, Morgan G,
Nielsen M, Barrett D, Smith G, Collins CH. Confidential inquiry into
quality of care before admission to intensive care. BMJ 1998; 316:1853-8.
5. Hourihan F, Bishop G, Hillman KM, Daffurn K, Lee A. The medical
emergency team: a new strategy to identify and intervene in high risk
patients. Clin Intensive Care 1995;6:269-72.
6. Bristow PJ, Hillman KM, Chey T et al. Rates of in-hospital arrests,
deaths and intensive care admissions: the effect of a medical emergency
team. Med J Aust 2000;173:236-240.
Competing interests: No competing interests
Editor,
The article by Vincent and colleagues is both salutary and helpful in
that it again draws attention to possible ways of improving practice.1 We
are concerned however that no mention has been made of what might be
called “no win” situations where there is a high chance of an adverse
event whichever course of action is pursued. As far as we can ascertain
from the “criteria for adverse events” no allowance is made for this and
indeed it is probably impossible to do so. However, it is important to
acknowledge such a problem when commenting on the findings.
Even in the example quoted there is an element of this dilemma.
Osteomyelitis as a complication of leg ulceration is probably quite rare,
whereas the alternative, apparently preferred option of more aggressive
management with antibiotics could have resulted in significant antibiotic-
induced complications.
Older people by virtue of the common problem of multiple pathologies
are particularly candidates for “adverse events”. Should you increase
treatment for cardiac failure in someone with a degree of renal impairment
and risk precipitating frank renal failure (adverse event of commission),
or risk them dying from undertreated cardiac failure (adverse event of
omission)?
Comparisons with civil aviation procedures seem popular in relation
to adverse events and risk reduction. Perhaps we could suggest “We are
flying a rather old and unreliable aircraft. Would you rather crash here
or there?”
David Griffith, Paul Diggory and Anand Mehta
Consultant Physicians - Care of Older People
1) Vincent C, Neale G and Woloshynowych M. Adverse events in British
hospitals: preliminary retrospective record review. BMJ 2001; 322: 517-9
(3 March)
No competing interests
Competing interests: No competing interests
The products of the database envisaged by the NHS adverse event pilot
study research team (BMJ 2001,322:517-519) would be very welcome. It would
be useful to couple the database with a convincing explanation for the
extent of adverse events and for why more resources have not been spent in
attempts to reduce either their extent or their adverse consequences,
despite the evident public concern.
Indeed, why has there apparently not yet been a systematic database that
encompasses adverse events across the NHS?
It might be that more of the scarce resources available to the NHS
would be directed to these purposes, by those in a position to do so, if
they could capture more of the benefits or savings from doing so or if
they themselves were to experience more adverse consequences associated
with the events in question. If so, attention to the relevant incentive
structures might prove to be part of a cost effective and appropriate
response to the issue.
In any case, the application of some additional insight into what
limits the commitment of scarce resources in such a way would complement
the prospective evidence base. Together, they might help in the
development of a workable strategy for improving the consistent quality of
service in the NHS for patients which their advocates seem to want. A
greater confidence in our ability to select a cost effective and
appropriate response, based on sound explanation as well as sound
information, might even increase the preparedness to bear the cost of a
better database more quickly.
Competing interests: No competing interests
Dear Sir,
At first sight the results are alarming.The rate of AEs for Australia-
16.6%,for the UK-10.8% and for the US-3.7%. We concede the importance of
measuring the extent of AEs and taking action to reduce their number.
However the measures used should be reliable and reproducible.It
should be noted that the mortality rates are in the reverse order, with
the US-13.6%, the UK-8% and Australia-4.9%. When charted the figures
provide an almost staight line.
The methodological problem most likely relates to the threshold
determination, with a high threshold for AEs providing small numbers of
serious events and a low threshold large numbers of minor events.This
study weakness is now recognised, but the community, particularly in
Australia and the UK need to be reassured that the health care standards
in their two countries are unlikely to be inferior(to the US) to the
extent suggested by these "comparative" studies.A large difference was in
the "eye" (or attitude)of the reviewers.
Competing interests: No competing interests
Professor Vincent and his team have produced some excellent, albeit
worrying work in this study. Many of us who work in the NHS may feel this
is the tip of the iceberg - particularly if adverse drug events (as
highlighted in another response) were not included.
It will be interesting to see how Government and Professional bodies
will respond. Clearly prompt action can be taken in some cases - I read
this morning in the 'Metro' that the Medicines Control Agency has imposed
a ban on Traditional Chinese Medicines that contain a substance that was
'blamed for causing kidney failure in two women'. Perhaps this is an
indication of double standards? That remains to be seen.
Competing interests: No competing interests
Re: Quick Risk Comparison
It seems to me that Barry Whittingham does not bear in mind that,
wherever you are (in a train, at home, or anywhere else), if you are
critically ill (it is unfortunately the case of many hospitalised
patients), you are much more likely to die than if you are healthy (it is
fortunately the case of most people who travel by train).
In other words: to compare trains with hospitals is quite risky,
scientifically speaking.
Wouldn’t it be more accurate to compare different hospitals (maybe
from different countries) with each other?
Competing interests: No competing interests