Individual at Bristol

A death in custody at
HMP Bristol in July 2004
Report by the Prisons and Probation Ombudsman
for England and Wales
May 2005
This is the report of an investigation into the circumstances of a
death in hospital on 4 July 2004 of a man who was a remand
prisoner at HMP Bristol. The man died from bacterial endocarditis.
A clinical review into the man’s treatment was carried out by Bristol
North Primary Care Trust.
We would like to extend our condolences to the man’s family for
their sad loss. I would like to thank the Governor in charge of
Bristol Prison, and his staff for their help.
Stephen Shaw CBE May 2005
Prisons and Probation Ombudsman
2
Contents
SUMMARY
INVESTIGATION PROCESS
HMP BRISTOL
EVENTS LEADING UP TO THE MAN'S DEATH
DISCUSSION WITH HEALTHCARE STAFF
BACTERIAL ENDOCARDITIS
AFTER THE MAN'S DEATH
LEVEL OF COMPLIANCE
FINDINGS
CONCLUSIONS
RECOMMENDATIONS
3
Summary
The man arrived at HMP Bristol on 29 June 2004. He was a remand
prisoner awaiting a court hearing to face a number of charges
relating to relatively minor offences. On arrival at Bristol, a
reception doctor examined the man and noted that he had a rapid
pulse which was attributed to drug withdrawal and he was taken into
the prison’s Healthcare unit.
On 30 June, test results for the presence of drugs in the man’s urine
proved negative and it was realised that his symptoms were due to
another cause. The man was transferred to outside hospital the
next day, but despite treatment he died on 4 July from bacterial
endocarditis. He was 33 years old at the time of his death.
The investigator spoke to one of the man’s sisters by telephone.
This report makes recommendations relating to the taking of clinical
observations of prisoners being monitored in the Healthcare unit.
Also the need to avoid delay in obtaining urine samples for testing
purposes, and the need to act promptly once urine test results are
available. However this investigation raises questions as to how far
an equivalent level of care can be offered by a prison’s Healthcare
unit compared to outside hospital.
4
Investigation process
My practice in cases of apparent deaths from natural causes is to
conduct an initial review to determine the extent of investigation
required.
My investigator visited HMP Bristol on 5 July 2004 when he was
given copies of relevant records, including the medical records. The
investigator spoke informally with the Deputy Governor and with a
prison chaplain who had visited the man in hospital, when she met
the man’s mother and one of his sisters.
The investigator spoke to the Chair of the Independent Monitoring
Board (IMB), who did not have any issues that he wished to draw to
our attention.
The investigator telephoned the man’s sister. She wanted to know
whether her brother had received adequate care and treatment.
She said that she particularly wanted to know whether there had
been any delay on the part of Healthcare staff at HMP Bristol in
recognising the seriousness of her brother’s clinical condition. Also,
once her brother was recognised to be seriously ill, she wanted to
know whether there been any delay in sending him into hospital.
Bristol North Primary Care Trust carried out the clinical review.
This report is based upon a thorough review of all relevant
paperwork, upon the clinical review and upon informal discussion
with Healthcare staff.
5
HMP Bristol
HMP Bristol is a local prison, which first opened in 1883. In recent
years the prison has been extensively refurbished and renovated.
In-patient beds in Bristol’s Healthcare unit are located within
individual cells, and are subject to the same level of security as
standard prison cells. At night-time, for reasons of prison security,
clinical observations of patients are not conducted: the term ‘clinical
observation’ means the measurement of several objective clinical
signs, for instance measurement of pulse rate, temperature and
blood pressure. Similarly, fluid balance charts are not maintained:
fluid balance charts are used to record fluids consumed, and urine
passed, by a patient. The only ‘observation’ of patients at night-
time, is a visual check made from the cell door observation hatch.
Only in the case of an emergency, will a Healthcare cell door be
opened at night.
6
Events Leading up to the Man’s Death
The man was received into Bristol in June 2004, when he underwent
a first reception health screen including examination by a reception
doctor. During screening, the man reported that he was an illicit
drug user: specifically he usually used heroin on a daily basis and
had last used this drug two days earlier. He also reported that he
used cocaine or crack cocaine on a daily basis and had last used this
drug a week earlier.
The reception doctor recorded that the man had tachycardia (a rapid
heart rate). Based on his symptoms, and on his reported use of
illicit drugs, the reception doctor concluded that the man was
suffering the effects of drug withdrawal and that he should be
monitored in Healthcare. The reception doctor made a note in the
man’s records that four-hourly observations should be carried out:
that is, that temperature, pulse rate, respiratory rate, and blood
pressure should be measured and recorded. The reception doctor
also noted that a fluid balance chart should be maintained.
On the morning of 30 June, the healthcare nurse made a note in the
man’s records: ‘Had a shower and exercise. Seen by [the
healthcare doctor]. No problems raised.’ Later that morning, the
healthcare nurse made a further note that the man had refused
lunch.
Also on 30 June, the healthcare doctor, made a note in the man’s
records to say that, due to what still seemed to be withdrawal
symptoms, he should remain under observation in the healthcare
unit and that this plan should be reviewed the following morning.
The man provided a urine sample at some time on 30 June and a
computer print-out, produced at 3.42pm, showed that the urine
sample proved to be negative for various drugs tested.
The man was seen by the detoxification team on the morning of 1
July. The record made of this review showed that the man had a
rapid heart rate, that he felt hot to the touch and that he was
slightly dehydrated. The record went on to say that, with these
continued apparent symptoms of withdrawal, but given the negative
urine result, the man needed to be seen by a doctor.
At 3.50pm on 1 July, the healthcare doctor reviewed the man and
noted his symptoms to be moderate to severe dehydration with a
feeling of weakness and lethargy. His observations showed him to
have a rapid heart rate, rapid pulse, rapid respiration and a high
temperature. The healthcare doctor noted that the man’s condition
had deteriorated compared to how it had been the previous day and
7
that he was also having difficulty holding down fluids. As the man’s
symptoms were clearly due to a cause other than drug withdrawal,
the healthcare doctor decided that he needed to be transferred to
outside hospital urgently. The man was taken to hospital, where he
was taken into the intensive care unit and later diagnosed with
bacterial endocarditis.
Over the following days, the man’s condition deteriorated further
and that led to him being transferred to another hospital just after
midday on 4 July 2004. The man’s condition continued to
deteriorate and that evening he was taken into the operating theatre
for surgery to be considered. The man died at around 9.30pm that
evening during surgery.
8
Discussion with healthcare and other Staff
The Senior Medical Officer (SMO) told my colleague that, when a
prisoner reports at first reception that he is a drug user, a urine
sample should be taken for testing. Sometimes a prisoner will be
unable to pass urine, perhaps because he is dehydrated or suffering
urinary retention. In such a case, the prisoner will be given water to
help him urinate.
The SMO explained that monitoring of a prisoner through four-
hourly clinical observations and maintenance of a fluid balance chart
was not possible overnight in a prison setting. Nor was such
monitoring very practical even during the day. The SMO said that,
if a prisoner’s condition is such that observations are necessary, it
would be more appropriate to send him into hospital rather than
keep him in Healthcare.
The reception doctor said that he had worked at HMP Bristol, in a
locum capacity, from February 2004 to August 2004 and had
worked day-time shifts only. The reception doctor said that when
he examined the man upon his arrival at Bristol on 29 June, he
concluded that the man was suffering the effects of drug
withdrawal. Ordinarily, prisoners suffering from withdrawal would
be allocated to a standard prison location, however the man seemed
to be having a more difficult time in withdrawal than usual, which is
why the reception doctor decided he should go to the healthcare
unit. The reception doctor did not know that the SMO did not
regard it as feasible for clinical observations and maintenance of
fluid balance charts to be carried out at night. The reception doctor
added that his role in the man’s care was only for preliminary
examination and preliminary planning. Once the man was in the
healthcare unit, it was for the doctors there to take forward his care
and treatment.
The healthcare doctor said that on 30 June, the man was weak, but
he was able to sit out of bed that day and had also been able to
collect his food. The man was drinking fluids and able to converse.
There were no indications that the man should be transferred to
hospital. On 1 July, the man was clearly worse than he had been
the day before and transfer to hospital was appropriate then.
The detoxification doctor saw the man in the clinic on 1 July. Ordinarily,
prisoners in drug withdrawal were seen by the detoxification team on
the morning immediately following their arrival into prison. They would
be reviewed in the clinic even if urinary analysis was still awaited. The
detoxification doctor could not, therefore, understand why the man was
not seen on 30 June. The detoxification clinic was held during the late
9
morning so the detoxification doctor surmised that he would have seen
the man at around 11am on 1 July. The detoxification doctor thought
that the man looked unwell and he mentioned his concern to the
healthcare doctor. The healthcare doctor said that healthcare would
keep the man under observation. This conversation would have been at
around midday.
The healthcare nurse was an agency nurse who worked at Bristol for a
brief period. The healthcare nurse remembered the man. She
remembered that he was dehydrated and was not drinking. She had
pushed him to drink and had made up Ribena drinks for him. The
healthcare nurse recalled the man being worse on 1 July than he had
been the day before. The healthcare doctor had asked her to take
observations and the man had been unable to support himself in bed
while she took those observations.
Other staff told my investigator that urine samples taken from
prisoners were stored in a fridge overnight and taken to the prison’s
Mandatory Drug Testing Unit (MDTU) at 8am the following morning.
Testing of the samples would normally be completed by about 9am.
Where a urine sample needs to be tested urgently, this can be done
on an ad hoc basis throughout the day. The MDTU does not open at
weekends, so a urine sample taken on Friday will not be tested until
the following Monday morning.
10
Bacterial Endocarditis
Bacterial endocarditis is an infection of the valves and inner lining of
the heart. It occurs when bacteria from the skin, mouth or
intestines enter the bloodstream and infect the heart valves and
lining. Although bacterial endocarditis can occur in anyone, people
with an abnormal heart valve or other heart defect are at greater
risk. Intravenous drug abuse is a risk factor, although there is no
evidence that the man was an intravenous drug user. People who
are immunosuppressed (those with a lowered resistance to disease)
are also at risk. Infective endocarditis frequently presents with
vague early symptoms which delays diagnosis. Infective
endocarditis can be treated successfully with antibiotics or surgery.
The average mortality rate for infective endocarditis is around 20
per cent, with higher rates applying when there are added
complications such as fungal infections.
After the man’s death
HMP Bristol followed its contingency plans for handling deaths
occurring at outside hospitals. A ward sister from the hospital
notified the man’s mother of her son’s death.
One of the chaplains from HMP Bristol had met the man’s mother
and one of her daughters when she visited the man while he was a
patient in hospital. Following the man’s death, the chaplain spoke
by telephone to the man’s mother and to both of her daughters.
The chaplain also attended the man’s funeral service, which was
held at the Baptist church to which the man’s mother belongs. The
chaplain told my investigator that there had been a great number of
mourners at what was a very moving service.
Level of compliance
The post-incident response by HMP Bristol was fully compliant with
Prison Service instructions and policies on managing a death in
custody.
11
Findings
At his first reception health screen on arrival at Bristol on 29 June,
the man reported that he was a user of both heroin and of
cocaine/crack cocaine. The reception doctor examined the man and
noted that he had a rapid heart rate. The reception doctor
concluded that the man was suffering the effects of drug withdrawal.
Normally, prisoners in withdrawal would be sent to a standard prison
location, but the reception doctor judged that the man was
experiencing a more difficult withdrawal than was typical. For this
reason, the reception doctor sent the man to the healthcare unit
where his condition could be monitored. The reception doctor noted
in the man’s medical record that four-hourly observations should be
conducted and a fluid balance chart maintained.
When examined by the healthcare doctor on 30 June, it would seem
that the man’s condition was largely unchanged from how it had
been the day before. The healthcare doctor recorded that his plan
was for the man to be reviewed the following day.
A urine sample had been taken for drug testing purposes at some
time on 30 June, and that afternoon the urine test result was
produced showing that the man was negative for various drugs
tested.
The man was reviewed by the detoxification doctor on the morning
of 1 July. The detoxification doctor made a note that the negative
result from the urine test, together with the fact that the man
remained unwell, meant that he needed to be reviewed by a
healthcare doctor. The detoxification doctor thought that the man
looked unwell and he mentioned his concern to the healthcare
doctor who said that healthcare would continue to keep the man
under observation. This conversation took place at around midday.
The healthcare doctor saw the man at 3.50pm and recorded that his
symptoms at that time included a rapid heart rate, rapid respiration
and that he was no longer able to hold down fluids. As these
symptoms could no longer be attributed to drug withdrawal, the
healthcare doctor arranged for the man to be transferred to hospital.
Initially, the man went to hospital and it was there that the
diagnosis of bacterial endocarditis was made. Despite treatment,
the man’s condition continued to deteriorate and on 4 July he was
transferred to another hospital. He died at the hospital later that
night.
12
As explained in the clinical review, the man died from a very rare
condition that is hard to diagnose. On admission to Bristol, his
presenting clinical symptoms were attributed to drug withdrawal and
that was not an unreasonable assumption to have made at the time.
Later, when it was recognised that the man’s symptoms were due to
some other cause, arrangements were made for his urgent referral
to hospital.
It had been the negative urine result that led clinical staff to
question the preliminary, initially reasonable, diagnosis that the man
was suffering from drug withdrawal. There was some delay in
obtaining this result. First, because the man was dehydrated he was
unable to produce a urine sample for somewhere in the region of 24
hours. That delay was compounded by a further delay before the
man was reviewed in the light of that result by the healthcare
doctor. Having said this, I recognise that the purpose of the urinary
analysis was for the purpose of devising a detoxification plan, and
not for diagnostic purposes. It is because of this distinction that it is
not considered necessary for the Mandatory Drug Testing Unit to
operate at weekends.
If urinary analysis is not used for diagnostic purposes, what
diagnostic tests or other objective measurements are used? The
reception doctor decided that the man should be kept in healthcare
where four-hourly clinical observations should be taken and a fluid
balance chart maintained. Clearly, the reception doctor thought that
such monitoring would be helpful in the management of the man’s
care and treatment. If he had known that such monitoring would
not take place, his approach to planning the man’s care might have
been different. The SMO has said that the taking of clinical
observations and maintenance of a fluid balance chart is impractical
in a prison setting. I accept that maintaining an accurate fluid
balance chart would be impossible without the prisoner’s total co-
operation. However, I reject any suggestion that the taking of
clinical observations is impractical; it would certainly not be
impractical during the day, even if it would be rather more difficult
to manage at night.
The healthcare doctor said that, when he examined the man on 30
June, his condition was not such that referral to hospital was
appropriate. It was on 1 July that the man’s condition had
deteriorated to the extent that his condition then, and in light of the
negative urine test result, indicated that he should be sent to
outside hospital. Had four-hourly observations been recorded, the
deterioration in the man’s condition might have been recognised at
an earlier stage, resulting in an earlier transfer to outside hospital.
13
It might have been the case, of course, that the man’s clinical
observations would not have indicated the need for his transfer to
outside hospital at an earlier stage. Indeed, the clinical review
concludes that “it is very unlikely” that the man’s condition would
have been picked up any earlier had he been in the community.
However, if four-hourly observations had been maintained, the
man’s records would have contained objective evidence to support
the clinical decisions made in his case.
14
Conclusions
When the man was received at HMP Bristol, his condition of bacterial
endocarditis was clearly already developing. Unfortunately, the
symptoms of this illness were assumed to be symptoms of drug
withdrawal. Given the man’s reported history of drug use, and
given that the diagnosis of bacterial endocarditis is not an easy one
to make, the assumption that he was suffering the effects of drug
withdrawal was not, initially, unreasonable.
There was a delay in obtaining a urine sample from the man, and a
further delay in the man being reviewed by a doctor once the urine
test result was available. Contrary to the advice of the reception
doctor, four-hourly observations were not taken and no fluid balance
chart maintained. That plan was probably not achievable in a prison
setting. It is also possible that even if an earlier, correct, diagnosis
had been made, the ultimate outcome for the man might have been
the same. Certainly, the clinical review concludes that the man was
appropriately looked after during his time in prison. However, the
absence of recording of clinical observations means that his care
was almost certainly not on a par with what he would have received
had he first presented with his symptoms at outside hospital.
15
Recommendations
I make the following recommendations, all of which are directed to
HMP Bristol’s Healthcare unit:
Local recommendation 1: All doctors at HMP Bristol – whether
locum or permanent – should be made aware of the frequency
and regularity of clinical observations that can reasonably be
undertaken within the healthcare unit. Where more frequent or
more detailed observations are indicated for a particular
prisoner, arrangements should be made to unlock the prisoner
during patrol periods, or to transfer the prisoner to outside
hospital.
Local recommendation 2: Where needed for testing, urine
samples should be obtained from prisoners at the earliest
opportunity. If a prisoner is unable to provide a sample, on
request, at reception, his care plan should include the aim to
obtain a sample with a minimum of delay.
Local recommendation 3: When a negative urine test result is
returned for prisoners being monitored in healthcare, the
prisoner should be reviewed by a doctor with an absolute
minimum of delay.
Local recommendation 4: The prison should consider the
purchase of voluntary drug testing analysis kits for use in
reception and healthcare. These will enable a urine sample to be
tested for the presence or absence of drugs within minutes of
production of the urine sample. This would prevent unnecessary
delay in diagnoses and treatment decisions.
16