SHERIFFDOM OF TAYSIDE, CENTRAL AND
Determination
by
ROBERT McCREADIE
Queen's Counsel, Sheriff of
Tayside, Central and
In the Inquiry into the
circumstances of the death of
GEORGE DONALD McLELLAN
In terms of section 6 of the Fatal
Accidents and Sudden
Deaths Inquiry (
The
Sheriff, having considered all the evidence adduced, DETERMINES as follows:
[1] In terms of section 6(1)(a) of the Fatal
Accidents and Sudden Deaths Inquiry (Scotland) Act 1976, that George Donald
McLellan (hereafter "the deceased"), born on 17 February 1968, and who
resided at 26 Milnab Street, Crieff, died between 0359 hours and
0440 hours on 24 December 2005 in Cell 4 at Divisional Police Headquarters,
Barrack Street, Perth.
[2] In terms of section 6(1)(b) of the said
Act, that the cause of death was Methadone and Dihydrocodeine toxicity combined
with chronic alcoholism.
[3] In terms of section 6(1)(b) of the said
Act, that the reasonable precautions whereby his death might have been avoided
were (a) to require confirmation of the vulnerability status of a prisoner
during a handover between duty officers;
(b) to carry out visits to vulnerable prisoners timeously.
[4] In terms of section 6(1)(e) the
other facts relevant to the circumstances of the death were (a) the monitoring
of persons in police custody by medically qualified personnel; (b) the absence of a Scotland-wide
system for the care of persons in police custody; (c) staffing levels during weekends; (d) the availability of life-saving drugs and
medical equipment in police stations;
(e) the clarity of the Tayside Standing Operating Procedure; and (f) the absence of a clear and simple "aide-memoire"
on prisoner care.
NOTE
Preliminary
[1] This inquiry was into the death in
police custody of George Donald McLellan, aged 37 years. Dr David Griffiths, the Procurator Fiscal,
appeared for the Crown, Mr Murphy appeared for Mr Frank McLellan, the
brother of the deceased, and Mr Reid appeared for the Chief Constable
of Tayside. On
[2] On 10 and 11 May 2006 I heard the
evidence of Dr David Sadler, who conducted the post mortem, and of Dr Paul
Guy, the police surgeon who examined the deceased on the evening prior to his
death. I also considered the evidence of
Detective Sergeant John Petrie, which was in the form of an affidavit, and
then heard from Police Constable Roderick Ritchie, Sergeant Kenneth
Copeland, Mr John Penman, Mr Alexander Innes and Sergeant Rodney
Wigley. Thereafter I adjourned the
inquiry until
[3] On
[4] That hearing took place on
[5] On
The Circumstances of Mr McLellan's Death
[6] What actually happened in this tragic
case was never seriously in dispute. At
0505 hours on
[7] Dr Sadler explained that the high
therapeutic level of dihydrocodeine, combined with the potentially lethal level
of methadone, created "a cocktail effect".
One effect of opiates is to depress the part of the brain stem called
the respiratory centre, which initiates each breath. As he graphically put it, what happens is
that sleep becomes coma, and coma becomes death. In the deceased's case the cocktail of drugs
depressed his respiratory centre, leading to respiratory arrest. Dr Sadler commented that the addition of
dihydrocodeine would further depress breathing, even in a tolerant user of
methadone. Professor Busuttil
agreed: in his opinion these two opiates
would have acted synergistically to depress breathing by a direct effect on the
respiratory centre, a state of affairs that could easily cause death even in a
person quite habituated and tolerant to the use of these drugs on a regular
basis. Dr Sadler considered that
the drugs could have been taken in the morning or afternoon of
23 December, before the deceased's arrest, and that this was entirely
consistent with his death in the early hours of the following morning.
[8] On
[9] The circumstances in which the deceased
came to be at Divisional Police Headquarters were explained by Police Constable
Roderick Ritchie. He was stationed at
Crieff Police Station, and knew the deceased.
At about 1925 hours on the evening of
[10] The deceased was then taken to Divisional
Police Headquarters in
[11] After giving his details at the Charge
Bar, and Sergeant Copeland going through the health requirements with him, the
deceased was searched by Mr Penman. He
found two blister packs (one empty and one with pills in it), a box of what
looked like prescribed drugs, and four small medicine bottles. Three of the bottles were empty and the
fourth had pills in it. The contents of
these various containers appear to have been propranalol, Vitamin B and
ibuprofen tablets. The deceased was then
removed to Cell 4. Because of what
was found on his person, he was thereafter subjected to a full strip search by
Constables Ritchie and McCraw, on the instruction of Sergeant
Copeland. This was a normal strip
search, involving an examination of all orifices. The officers detected no sign that the
deceased's life might be in danger, and found no drugs on his person.
[12] Sergeant Copeland had been a Charge Bar
officer on and off since 1996, and was also involved in training staff in how
to deal with vulnerable prisoners. He
explained that the first stage of processing a prisoner was to carry out a
"vulnerability assessment". This had two
aspects: the vulnerability of the
prisoner and the risk to the officers.
In the deceased's case there was no risk whatsoever, and the focus was
accordingly on his vulnerability. The
assessment involved examining the deceased to address his care and welfare
needs. Like Constable Ritchie, Sergeant
Copeland knew the deceased from previous dealings with him, and he was aware
that the deceased had "issues" with both drink and drugs.
[13] Sergeant Copeland began the assessment by
reading out a statement to the deceased.
This informed him that he would be asked questions about his care and
welfare whilst in custody, and that it was important to answer them accurately
and honestly. This statement was
contained in the Admin 51 Custody Record form (hereafter "the Custody
Record"), which recorded relevant information about his vulnerability. One of the questions that
Sergeant Copeland asked the deceased was whether he was taking any
medication. The deceased mentioned
diazepam and amitryptilene, which were both noted, but he did not mention methadone. However Sergeant Copeland added methadone to
the Custody Record after opening one of the pill bottles and finding a green
syrupy residue on the lid, which the deceased told him was methadone.
[14] The deceased was then asked if he took any
other kind of drugs, and he referred to cannabis. Sergeant Copeland wrote this down, and later
added heroin on being told by the deceased that he took methadone for his
heroin addiction. Sergeant Copeland also
noted the deceased's admission that he was an alcoholic. The deceased was then asked if he had used
any drugs or consumed any alcohol within the last 24 hours, and he denied doing
so. Sergeant Copeland said that he had
dealt with him on several previous occasions, when he had been both drunken and
incoherent, but on this occasion he considered him to be "the best and
soberest" that he had ever seen him.
Accordingly he believed the deceased's denial.
[15] The deceased's "vulnerability" status
required to be entered on the Custody Record, and there were only two
options; to mark him as highly
vulnerable (hereafter "HV) or as non-vulnerable (hereafter "NV). Notwithstanding the deceased's denial,
Sergeant Copeland marked the deceased as HV, and placed him on half-hourly
observations. The reason that he gave in
court was clear and succinct, and merits reproduction: "In my experience any individual who abuses
drugs, takes medication and is an alcoholic is a vulnerable person in custody,
either from withdrawal or for any other reason.
I would declare that person as being at high risk unless satisfied there
was nothing untoward". In addition, and
because he was worried that the deceased might have taken drugs, he requested
the attendance of a police surgeon to examine the deceased.
[16] Dr Paul Guy, a General Practitioner and
police surgeon of some nine years' standing, examined the deceased at around
2100 hours. He was aware of the drugs
recorded by Sergeant Copeland, and described them as "a fair mix". He found the deceased unsteady on his feet,
and vague about what he had taken. In
his medical notes he wrote: "obviously been drinking". He also found the deceased's pupils to be 3mm
in diameter, and unreactive - a sign that he was taking opiates, although not
specifically heroin. The deceased's
pulse was regular, and his blood pressure and respiratory rate were
normal. As his speech was a little
slurred, and he was slightly incoherent, Dr Guy gave him a score of 14 out of
15 on the Glasgow Coma Scale. Because he
was unfamiliar with the deceased, and had no idea whether he was going to get
better or worse, Dr Guy then decided to place him on quarter-hourly
observations. He noted in his police
surgeon's log and in the Custody Record that these should revert to half-hourly
observations after a couple of hours.
[17] It was Mr Penman who took the deceased to
the examination room to see Dr Guy.
He thought the deceased's condition had deteriorated somewhat since his
arrival at the Charge Bar: he was more
unsteady and vague, and there was less opportunity for communication or
rapport. Mr Penman was present at Dr
Guy's examination, and observed that the deceased looked quite vague when asked
if he had taken anything. He recalled
that the deceased was unco-operative and would not answer Dr Guy's
questions. Mr Penman was not asked if he
told Dr Guy about his perception of deterioration, but I should say at this
point that there is nothing to suggest that the absence of such information
made any difference to Dr Guy's clinical assessment, or to the manner in which
the deceased was subsequently handled.
Dr Guy was careful and cautious in his examination, and reflected that
in his instructions about the frequency of monitoring. Furthermore, the deceased was no worse when
Dr Guy saw him in his cell an hour later (see paragraph 23 below).
[18] Mr Penman told Sergeant Copeland about Dr
Guy's instructions, and at 2130 hours the sergeant recorded the change to
quarter-hourly observations in the deceased's Custody Record. He also decided to retain the deceased's HV
status, which he said reflected the advice given by Dr Guy. Sergeant Copeland then remained on duty until
the end of his shift at around 2300 hours, when he handed over to the incoming
duty officer, Sergeant Rodney Wigley.
Between 2105 hours and 2215 hours the deceased was observed in his cell
at regular quarter-hourly intervals by Mr John Penman, the duty custody care
assistant. These visits were recorded on
another document, Form Admin 51(a) (hereafter "the Cell Sheet"). The Cell Sheet recorded the date and time of
visits to a prisoner, the type and dosage of any medication, and remarks about
each visit. At the top of the "remarks"
column was the following: "A verbal response
must be obtained on every visit". In the
course of the inquiry it became clear that such a response is no longer
required from NV prisoners.
[19] At 2115 hours Mr Penman recorded in the
Cell Sheet: "Standing at door"; at
2130 hours he recorded: "At door stated OK in order"; and at 2145 hours, 2200 hours and
2215 hours he recorded: "Lying on floor woken up". Sergeant Copeland explained that it was very
common for prisoners to lie on the floor rather than on the bed, and I accept
that there was nothing untoward about this.
Mr Penman confirmed that he went into the cell every quarter-hour to
obtain a response from the deceased. He
described his role as one of monitoring rather than simply observing, and said
that he was getting a response from the deceased either verbally or by movement. As he put it: "In my experience, if someone
is under the influence of drink, they go into a deep sleep to sleep it
off. You can stand at the door and
shout, but get no response. I go in,
shake them, get a response and am happy with that". He said that was what he normally did with
quarter-hourly observations - he did not let a visit go by without waking up a
prisoner.
[20] Just before handing over to Sergeant
Wigley at 2255 hours, Sergeant Copeland recorded in the Cell Sheet that the
deceased was receiving no medication, and was to be kept on warrant. The handover then took place. The procedure involved the outgoing duty
sergeant discussing each of the prisoners in the charge room, and referring to
the on-screen digital record, the nature of the offence, and any relevant
issues, including the question of vulnerability. Sergeant Copeland confirmed that the
procedure was followed on this occasion - he discussed every prisoner in
custody with Sergeant Wigley, and spoke to him on two occasions about the
deceased. A physical check was also carried
out in the cell passage. Sergeant
Copeland said that it was not a requirement for the incoming duty sergeant to
see each of the prisoners, but that was what happened on this occasion.
[21] The next entry in the Cell Sheet was made
by Sergeant Wigley, at 2301 hours, in Sergeant Copeland's presence. It referred to the deceased snoring, and
recorded the change to half-hourly visits, as instructed by Dr Guy. Sergeant Copeland said he and
Sergeant Wigley discussed this particular change, which was consistent
with Dr Guy's advice, and he was quite happy with it. He was satisfied that there was no change in
the deceased - he was stable, and had not worsened. However, he did not actually see what
Sergeant Wigley wrote on the Custody Record - he only recollected Sergeant
Wigley saying that he would move the deceased to half hourly visits. Nor was he aware that Sergeant Wigley had
changed the deceased's status from HV to NV.
He said that he did not consider it unusual for a change in status not
to be mentioned. Thereafter
Sergeant Copeland went off duty. On
reflection he believed that there was nothing he would have done differently,
and he expressed considerable surprise at the fate of the deceased.
[22] Importantly, it was Sergeant Copeland's
opinion that half-hourly checks meant the retention of the deceased's HV
status. However, he told the inquiry
that the opinion he expressed when discussing the deceased with Sergeant Wigley
was that his vulnerability was greatly reduced, and he could be monitored with
hourly checks. Sergeant Copeland said
that he would have been happy to go for hourly checks, and make the deceased a
NV prisoner, but Dr Guy's instruction for half-hourly checks took precedence. That meant the retention of HV status. As he put it, the vulnerability decided the
checks. He agreed during
cross-examination that ascribing NV status to a prisoner, whilst making
half-hourly checks suggestive of HV status, sent out a mixed message - someone
looking at the NV classification might be unsure about whether or not to rouse
the prisoner from his sleep, and the forms that were available did not assist
in clarifying what should be done.
[23] Mr Penman was not the only person who
visited the deceased before Sergeant Wigley took over at 2300 hours. Dr Guy saw the deceased at about
2200 hours, when he was passing his cell to see other prisoners. He went into the cell, found him lying on the
floor (which is consistent with Mr Penman's observations) and asked him to get
up. The deceased got up, and wandered
round the cell. Dr Guy considered that
he was no worse than when he examined him:
his demeanour was the same, as was his speech, which was slurred. Dr Guy agreed that he was "obviously a cause
for concern", otherwise he would not have gone into his cell, but no more so
than before.
[24] Sergeant Wigley, who had 26 years' police
service, said that the formal handover from Sergeant Copeland was at 2300
hours, but he normally came in about half an hour earlier to facilitate
it. He accepted that he was the person
ultimately responsible for the classification of prisoners as HV or NV, but
said he would be guided by a police surgeon, whose decision would be
"gospel". He confirmed the discussion
between himself and Sergeant Copeland.
He explained that on going into the cell passageway to start the
handover, Sergeant Copeland was a couple of prisoners ahead of him. On looking at the Cell Sheet he was unable to
read Sergeant Copeland's writing, and asked him for an explanation of the
deceased's circumstances. They then
stood for two or three minutes while Sergeant Copeland briefed him. Thereafter he entered the half-hourly visits
in the Custody Record. Sergeant Wigley
remembered that Sergeant Copeland was present when he did this, but he did not
recall whether he was actually standing beside him.
[25] Sergeant Wigley was asked why he altered
the deceased's vulnerability status from NV to HV, at the same time as he
entered the half-hourly visits. His
reply was concise, and to the point: "It
is incorrect. The 'non' should be
'high'. The rest of it is correct". He agreed that he had written "non" instead
of "high", but he was unable to explain his error. He said that his training told him that
someone on half-hourly checks was vulnerable - if he had seen the words
"non-vulnerable" and "half-hourly" together, he would have been aware that it
required to be changed. When pressed
further by Dr Griffiths on why he therefore failed to spot the error, he said
this:
"I am not a great fan of quarter hourly checks. I use them very frequently. If there is that concern about someone, they
should be on constant observation. My
explanation is seeing this person is on quarter hourly I have decided to change
them to half-hourly and discussed [that] at great length with Sergeant
Copeland. I am saying this situation is
improving, changing from quarter hourly to half hourly, that would make him
'non'. I don't have a particular explanation. But my contention would be that that has not
made a particular difference. Someone
on half hourly checks is being treated as highly vulnerable [my emphasis]"
When Mr Murphy asked Sergeant
Wigley why he had written down "NV", his reply on this occasion was: "For some reason, whether distracted
speaking, I don't know".
[26] Sergeant Wigley acknowledged the possibility that someone
seeing the NV classification might think there was no need to obtain a verbal
response, notwithstanding the half-hourly visits. He was then referred to the entries in the
Cell Sheet made by Mr Alexander Innes, his custody care assistant, between
2335 hours and 0345 hours, and these were contrasted with Mr Penman's
earlier entries. It was suggested to him
that whereas Mr Penman's entries showed a pattern of quarter-hourly visits,
with the deceased being wakened up, Mr Innes' entries - at 2335 hours
("asleep in order"); 0005 hours
("okay in order"); 0029 hours and
0102 hours ("asleep in order");
0135 hours ("okay in order");
0204 hours ("okay responded");
0235 hours ("asleep in order");
0306 hours ("lying on floor woken up") and 0345 hours ("asleep in
order") - appeared to indicate that he was dealing with the deceased as a NV
prisoner. Sergeant Wigley did not know
whether that was so, but he did not disagree.
[27] Mr Alexander Innes then gave evidence. He began work as a custody care assistant in
June 2004 after spending 32 years as a prison officer. Like Sergeant Wigley, he came in at around
2230 hours, and there was then a general handover between himself and
Mr Penman. Although Mr Penman told
him who the deceased was, and that he had been going into his cell to obtain a
response from him, he did not remember being told that the deceased was under
the influence of drugs. Mr Innes made
his first entry in the Cell Sheet at 2235 hours, and explained that
this involved going into his cell with Mr Penman, who gave the deceased a
light shake. There was a response,
although he could not recall what is was, and they were both satisfied with
him. His remark in the Cell Sheet on
that occasion was "Okay in order".
[28] Mr Innes said that the Custody Record was pinned up beside the
door to the cell, and was obvious to anyone entering it. He told the inquiry that he had indeed
treated the deceased as NV, because that was what Sergeant Wigley had written
on the Custody Record. He was quite
happy with him at 2335 hours, when he noted:
"Asleep in order". He was also
satisfied with him at 0005 hours, when he noted: "OK in order"; at 0029 hours and 0102 hours, when
he noted: "Asleep in order"; and at 0135 hours when he noted: "OK in
order". However at 0204 hours he
noted: "Okay responded". He explained that the deceased's response on
this occasion was not verbal, but "more like a grunt". At 0235 hours he recorded: "asleep in order"; and then at 0306 hours he noted: "Lying on
floor woken up". Mr Innes said that he
went into the cell on this particular visit because the deceased had moved
position. He had previously been half
sitting on the floor and leaning on the bed, but on this occasion he was lying
between the corner of the wall and the bed area. Wondering why he had moved, he went in and
roused him, probably by shouting his name a few times. This time the deceased responded with a
grunt, or what Mr Innes described as an "ahah".
Like Sergeant Copeland, Mr Innes remarked that it was not unusual for
prisoners to lie on the floor, and he remained quite happy with the deceased's
condition.
[29] At this point it is necessary to refer to the affidavit
evidence of Detective Sergeant Petrie. In
the course of his investigation Sergeant Petrie established that the times of
two of the visits recorded on the Cell Sheet by Mr Innes were inconsistent with
what was shown on the CCTV system, which was recording the time
accurately. In particular, the recorded
visit at 0345 hours appeared to have taken place at around 0359 hours, and the
recorded visit at 0420 hours appeared not to have happened at all. Sergeant Petrie's evidence was not disputed
by any of the parties. It was therefore
put to Mr Innes that his next visit, at 0345 hours - when he wrote:
"Asleep in order" - probably took place at around 0359 hours, some
53 minutes after the previous visit at 0306 hours. Mr Innes said that he could not really
assist, but it was quite a busy night, and he might have been doing something
else. Since he was the only custody care
officer on duty, and more prisoners than usual were released that night because
of the Christmas period, he believed that he might have inserted the wrong
time.
[30] Mr Innes explained that on this occasion he did not actually go
into the cell, which he would have done if the deceased had been classified as
an HV prisoner. The deceased remained on
the floor, in roughly the same position shown in the photographs taken after
his death, although more on his side, and with his right shoulder more
upright. Notwithstanding the fact that
the deceased was facing away from the door hatch, and his face was not visible,
Mr Innes was able to see a chest movement.
He was therefore satisfied that the deceased was breathing, and also
content that he remained asleep.
[31] Mr Innes agreed that there was no visit at 0420 hours, but he
could not explain how the mistaken entry came about. He accepted that the next visit actually took
place at 0440 hours, which was some 40 minutes after the previous
one. He explained that whether or not it
was possible to do half-hourly visits "spot on" depended on the
circumstances. Although he tried to keep
as near as possible to the half-hourly intervals, it could be affected by
having to assist with a search, or help with fingerprinting, or photograph a
prisoner. When asked, he also said that
he had never been told to prioritise a vulnerable prisoner, although he believed
that should be the practice.
[32] On this occasion Mr Innes looked through the hatch, realised
something was wrong, and entered the cell.
He then went straight through to the front office and shouted on
Constable Murray, who was the constable on duty, to come through. Constable Murray and Sergeant Wigley
both came, and they re-entered the cell.
He recalled Constable Murray saying: "I think he's dead". Dr Guy was then called, and he arrived
at 0505 hours. Dr Guy
explained in his evidence that he found the deceased lying on the floor of the
cell, in the three-quarters prone position shown in the photographs. He noted that there was no pulse, there were
no breath sounds, and the deceased's pupils were fixed. Consequently, he pronounced him dead. Dr Guy said that he could find no signs of injury,
only mucous and what appeared to be vomit.
[33] Sergeant Wigley was unaware at the time that no visit had taken
place between 0306 hours and 0359 hours. However, like Mr Innes, he explained that it
was the Friday night before Christmas, the next custody court was not until the
Tuesday or Wednesday of the following week, and there had consequently been a "great
rotation" of prisoners. His examination
of the compute records revealed that he had released nine prisoners that
evening, including one at 0328 hours and another at 0335 hours. Since Mr Innes had brought those prisoners to
the Charge Bar, he thought this might explain the time gap. Sergeant Wigley was also unaware of the
mistaken entry at 0420 hours, which he accepted was incorrect. When Mr Murphy put it to him that 40 minutes
had therefore elapsed since the previous visit, and the 10-minute delay might
be important, Sergeant Wigley accepted that it might be. However he emphasised that it was an
imperfect world, and delays could occur during busy periods. He then made the comment that in the case of
the deceased, rigor mortis had obviously set in, and therefore the 10 minutes
would not have been vital.
[34] The Tayside SOP requires the duty sergeant to visit all
prisoners, where practicable, at least three times during the shift. Sergeant Wigley observed that this did not
require him to visit the prisoners at any specific time. His only visit to the deceased that evening
had been at the handover with Sergeant Copeland, and he felt that if he had
checked the deceased again himself, "something might have rung an alarm
bell". However, he had been unable to
do that because it was such a busy evening - at 0440 hours he had just sat down
to complete the review of prisoners in custody, although he would probably have
carried out a further visit within the next few minutes. For the avoidance of doubt, I do not consider
that Sergeant Wigley can be faulted in this matter. It is indeed correct that the visits by the
duty sergeant do not need to be at any particular time. Furthermore, as Chief Inspector Tonks
remarked, the visits tend in practice to take place at the beginning, middle
and end of the shift.
[35] I now turn to the evidence about the "distinct verbal response"
that the Tayside SOP states should be obtained from HV prisoners. All the police witnesses on duty that evening
spoke about their understanding of what was required. Mr Penman said that if half-hourly visits
were instructed, he expected a response ever alternate visit - meaning every
hour - with the response being obtained by going into the cell and shaking the
prisoner. That was something he was told
to do, and he agreed that it was "policy".
He thought everyone knew that a response had to be obtained for
half-hourly and quarter-hourly checks.
He was aware that a verbal response was not required for hourly visits,
and that the Cell Sheet did not reflect what happened in practice. When asked to comment on Mr Innes'
entries between 2335 hours and 0135 hours, which stated "asleep in
order" and "okay in order".
Mr Penman said that if the prisoner was snoring, "I would take it
he was okay". He was asked if that was a
verbal response, and replied that the prisoner was deemed okay because he was
snoring. He added that if he shouted a
prisoner's name, and the prisoner put his hand up, he would take that as a
verbal response.
[36] According to Sergeant Copeland the guidance suggested that a
non-vulnerable individual should be checked hourly, and there was no need for a
verbal response. However, if a prisoner
was classified as highly vulnerable, the visits would be at least half hourly,
if not more frequent, and required responses from the prisoner. Sergeant Copeland said this meant that
the prisoner should be disturbed - some verbal response was what he would look
for. For those classed as non-vulnerable
it was different: in such cases it was
sufficient if a prisoner could be heard breathing, or snoring. Sergeant Copeland did no explain exactly
what he meant by a verbal response.
[37] When Sergeant Wigley was asked about this matter, he said that
on each visit to a HV prisoner, the officer concerned should obtain "a distinct
verbal response" as per the standard operating procedure. However, that was sometimes difficult to obtain,
and he himself frequently wrote "uhuh" or indicated a grunt. Some prisoners signalled with a foot or an
arm, which was possibly sufficient to meet the requirements of a verbal
response, but they were usually NV prisoners.
Mr Innes' position was that the correct way to check a HV prisoner was
quarter-hourly, with a response from the prisoner, but this did not always need
to be verbal. There were various ways in
which a prisoner could respond to his name being called, including the movement
of a hand or a foot. If a HV prisoner
appeared to be asleep, and there was no response from the hatch, he tended to
go into the cell and give him a shake.
When I asked him whether he required a verbal response from every HV
prisoner at each visit, he replied: "If
I get any response from them I will take it as a response".
[38] After these witnesses had completed their evidence I was
informed by Mr Reid that there were no national guidelines for
[39] A copy of the Tayside SOP in force at the time of the
deceased's death was thereafter produced.
A copy of the guidance produced by the National Centre for Policing
Excellence on behalf of the Association of Chief Police Officers and the Home
Office - "Guidance on the Safer Detention & Handling of Persons in Police
Custody 2006" - was also made available (hereafter "the NCPE Guidance"). Evidence on the Tayside SOP was then given by
Chief Inspector David Tonks, who told me that he chaired the Tayside force's
Custody User Group, which brings together management and practitioners from
around the force to consider matters of custody.
[40] Chief Inspector Tonks commented first of all on the following
passage from the Tayside SOP (page 9):
"The utmost care is to be exercised when dealing with persons who are
found unconscious, semi-conscious or are suspected of having taken drugs. In such cases, or where there is doubt, the
prisoner should in the first instance be taken directly to a hospital. All such information about a prisoner must be
disclosed by arresting/presenting officers, to allow an accurate assessment to
be made for their care at hospital and during custody".
The Chief Inspector said that
this served to focus the attention of officers on the risks associated with
drug abuse, and on the need for a reasonable assessment. A person would be taken to hospital if there
were acute signs of distress (such as a low level of consciousness), suggesting
that something might have been taken immediately beforehand and there was
therefore likely to be a significant risk to the individual. As he put it, the brief was to play it safe.
[41] I note that the Tayside SOP contains the following specific
provisions about the maintenance of a Custody Record (page 13):
"A prisoner custody record is to be raised and
maintained in respect of every prisoner who is to be lodged in a cell or a
detention room. Accurate completion of
this record incorporating all information and actions relating to that prisoner
is essentially and the officers presenting the prisoner must be asked
specifically whether they have given all available information which may bear
the vulnerability assessment which has to be made for that prisoner .....
"Custody officers on finishing duty should brief the
oncoming supervisor on the status of all prisoners. The oncoming supervisor taking duty should as
soon as practicable, visit all prisoners, check their status and update the
cell sheet accordingly .....
"During the time a prisoner is in custody, the ADM.51
Custody Record is to be displayed beside the cell door of the prisoner and
after the prisoner's release, be retained in chronological order at the
station. ......."
[42] There then follows a passage on the assessment of vulnerability
(pages 13-14). It is as follows (the
paragraph in italics is to be amended, as I later explain):
"Immediately after the reception of a prisoner the Officer in Charge
[my emphasis] is to carry out an assessment of the VULNERABILITY of the
prisoner based on all available information.
The assessment will take account of factors such as:
(a) age or general frailty;
Remember that all children will be regarded as "high vulnerability"
(b) state of sobriety;
(c) known medical condition;
(d) mental disposition;
(e) self harm/suicide risk, including offence impact;
(f) known drug dependency [my emphasis].
"Where it appears that a
prisoner is mentally disordered a doctor is to be informed and requested to
attend.
"Prisoners who are drunk and
incapable, under the influence of drugs, potential suicides, mentally
disordered or suffer known potentially serious medical conditions such as
epilepsy, or have been exposed to CS Spray, will automatically be considered
high vulnerability. In addition, if any
advice is given by a Doctor on these issues then this is to be used in
assessing the frequency of checks on such a prisoner.
"If a prisoner is placed in the high vulnerability category the reasons
for this assessment will be entered in the space provided on the Adm 51
prisoner custody record in red ink. In
addition, the nationally agreed form,Tayside-Adm.19, will be completed and
attached to the Adm 51 prisoner custody record.
"If there is a change to the
vulnerability assessment of a prisoner while in custody this will be entered in
red ink in the remarks column and added to the box on the front of the
Adm 51 Prisoner Custody Record at the time any change is made, with
corresponding adjustment of frequency of visits. The original entry should be lightly crossed
through, such that it remains on record".
[43] The Tayside SOP also includes the following
guidance (at page 21) on visits to prisoners by staff (the passages in italics
are due for amendment):
".... The frequency of visits is
to be determined by the Officer in Charge [my emphasis] who must specify
on the custody record the minimum intervals between visits to all prisoners in
the high vulnerability category which may range from constant observations
to intervals of up to 30 minutes [my emphasis]. Initially such intervals
are to be specified on the custody record.
Any changes to the specified intervals or to the vulnerability category
are also to be shown in red ink, on both the timed entry and the front of the
Adm.51 form. Prisoners not assessed as
high vulnerability category are to be visited at least once every hour.
"A distinct response is to be obtained from every prisoner that is
classed as high/vulnerable. This is to
be carried out on every visit. The words
spoken by them will be recorded in the remarks column on the custody record
form ADM 51 along with any other features such as requests made by the
prisoner.
"Prisoners who are classed as non-vulnerable should be visited hourly but
there is no requirement to obtain a verbal response ........
"Officers in Charge must
immediately upon taking duty, visit every prisoner and familiarise themselves with
the full contents of the custody records to determine whether good reason to
continue custody exists and ensure that adequate levels of supervision continue
to be applied. At this time they are to
ensure that any support staff involved with prisoner care are fully briefed as
to their specific duties in respect of each prisoner. They must attend the cell area where
practicable on at least three occasions during their tour of duty and must
visit all prisoners.
"Custody Officers and Custody
Care Assistants being relieved or going off duty are each to ensure that their
respective incoming relief is fully appraised about every prisoner who is
vulnerable, or special risk or on medication.
Each officer is to sign the custody record to this effect." ........
[44] Chief Inspector Tonks considered that Sergeant
Copeland's actions were in accordance with the Tayside SOP. However, during cross-examination by
Mr Murphy, he did acknowledge that although the Tayside SOP required a
distinct verbal response from every HV prisoner, none had been recorded from the
deceased in the period before 2300 hours.
He considered that the entries at 2145 hours, 2200 hours and
2215 hours should have included a distinct verbal response, which would
have been seen by the duty sergeant, and that the absence of such a response was
contrary to the Tayside SOP. Since a
distinct verbal response was a test of the threshold level of consciousness, it
was essential that such a response should be obtained from a HV prisoner. Nevertheless, although he accepted that this
was not "best practice", he commented that Mr Penman had found "other
circumstances equivalent to a distinct verbal response".
[45] The Chief Inspector considered it "most
unusual" for there to be half-hourly checks with a NV prisoner, who was
typically on hourly checks. He could not
remember any NV prisoner on a half-hourly check, and agreed that if he saw a
reference to such a period, he would assume that he was dealing with a HV
prisoner. He added that although such an
unusual entry stood out to him, and he hoped that might be picked up by a
diligent custody care assistant, he was sufficiently realistic to appreciate
that this might not happen. It was not
something that was explicitly clarified anywhere.
[46] Chief Inspector Tonks told the inquiry that
the requirement for a verbal response had been restricted to HV prisoners in
2004, and he was unable to say why the Cell Sheet had not been altered to
reflect this. The change had been made
because of concern that it was disproportionate - and perhaps contrary to human
rights - for NV prisoners to be wakened up at regular intervals throughout the
night. However, although this meant that
there was no need to obtain a verbal response from a NV prisoner who was
sleeping, or apparently sleeping, it was still entirely appropriate to rouse
him if a visual check suggested that something might be amiss - for example, if
he had not moved, or was snoring.
Officers were instructed to react to such non-verbal indicators on a
case by case basis.
[47] Chief Inspector Tonks confirmed that during
a handover it was normal procedure for a meeting to take place between both
duty sergeants, typically before visiting the cells. Any issues of note were discussed then, and
both sergeants walked around the cells together, discussing the regime for each
prisoner. The Chief Inspector said that
changing the vulnerability assessment depended upon a sergeant's own judgement
and any particular guidance from the police surgeon. If a sergeant felt that it was appropriate
to reduce the frequency of visits, he would make the necessary change. He agreed that there was no set procedure
which required the outgoing sergeant to see any change made on the Custody
Record by an incoming sergeant, pointing out that changes might be made by
individual sergeants half way through a shift, without the involvement of
anyone else.
[48] Chief Inspector Tonks accepted that there
was no procedure in place to ensure that Sergeant Wigley's error did not happen
again, but he emphasised that he was unaware of any similar error having been
made in the past. He saw no need for an
incoming custody sergeant to cross-refer to anyone else, including the outgoing
sergeant - the incoming officer was the sole decision maker, and took action
himself on the basis of the information and guidance he received. Nor did he consider cross-referring between
sergeants to be a viable proposition, given the annual number of prisoners in
Tayside (20,000), the frequency with which the status of prisoners changed, and
the fact that it would require the presence of two sergeants - with perhaps a
third sergeant to arbitrate in any dispute.
However the Chief Inspector did accept that there were normally three
handovers a day, involving sergeants being together for about 15 minutes during
each handover, and it was therefore feasible for a cross-check to be done at
that time, with the outgoing sergeant countersigning the change in status.
[49] The Chief Inspector was satisfied that the
entries in the Cell Sheet from 2335 hours to 0306 hours represented a
fairly typical pattern for a non-vulnerable prisoner, and nothing struck him as
out of the ordinary. However, he found
the gaps between 0306 hours and 0359 hours, and between
0359 hours and 0440 hours, surprising for a prisoner on half-hourly
observations. It was not unknown for
such a period to occur, but it was not the norm. Tayside Police had made an assessment of the
number of staff necessary to maintain the regime of checks, and typically they
had additional staff at weekends. In
[50] Chief Inspector Tonks agreed with Dr Griffiths
that the deceased had fallen between two stools; although subjected to a regime of checks more
appropriate to an HV prisoner, he had mistakenly been classified as a NV
prisoner, with the consequence that he had failed to receive the required level
of attention. Had the deceased retained
his HV status, Chief Inspector Tonks said that he would then have expected
to see a distinct verbal response recorded - such as "yes I am OK" - which
would have been impossible to elicit without going into his cell. The Chief Inspector told me that what he
meant by a "distinct verbal response" were actual words - a grunt or a raised
hand was not sufficient.
[51] Chief Inspector Tonks then gave evidence
about the proposed changes to the Tayside SOP.
He said that there was no concern about the adequacy of the Tayside SOP
itself, since the mistaken classification was a deviation from the
guidance that it contained. Sergeant
Wigley ought to have recorded the deceased as highly vulnerable, but even if he
had, the deceased could still have died.
That said, the Tayside SOP was recognised as a living document which
took account not only of fatalities and "near-misses", but also of good
practice elsewhere. Draft changes were
now proposed, and once the various stakeholders had been consulted these would
be presented to the force's executive for ratification. Although the changes were not finalised,
there was every indication that they would be accepted.
[52] Chief Inspector Tonks produced a draft of
the proposed changes and took me through them.
It is proposed to replace a paragraph in the section on vulnerability
(see paragraph 42 above) with the following:
"Following the vulnerability
assessment and taking into account all other known information, the officer in
charge is to make a decision as to whether a prisoner falls into the HIGH or
LOW vulnerability category.
"Prisoners of HIGH vulnerability
are to be checked at intervals of not more than half an hour but may be subject
to more frequent checks or constant observations, depending on the extent and
level of their vulnerability.
"Prisoners of LOW vulnerability
are to be checked at intervals of not more than one hour.
"The officer in charge will
record the vulnerability category and the frequency of checks in the appropriate
sections on page 1 of the Adm 51; these
entries to be made in red ink.
"The reason(s) for placing a
prisoner in the HIGH vulnerability section are to be recorded in the
appropriate section on page 2 of the Adm 51".
[53) The reason for this change is the Custody
User Group's view that classifying prisoners as NV or HV does not reflect
ordinary experience, and arguably gives out a misleading message that there are
prisoners who do not require to be cared for.
Chief Inspector Tonks said that the proposed change highlighted to
all responsible for custody care that every prisoner might in some measure be
vulnerable. It meant that there would
henceforth be two categories of vulnerability:
high vulnerability prisoners and low vulnerability ("
[54] Chief Inspector Tonks drew the
inquiry's attention to a proposed change in the procedure for visits to
prisoners. The intention is to
incorporate into the Tayside SOP the "Rousing Procedure" contained in the NCPE
Guidance for prisoners from whom a distinct verbal response cannot be
obtained. The proposed amendment (which
replaces the section in italics at paragraph 43 above) is as follows:
"A distinct verbal response is to
be obtained from every prisoner that is classed as high vulnerability
category. This is to be carried out on
every visit. The words spoken by them
will be recorded in the remarks column on the custody record form ADM 51 along
with any other features such as requests made by the prisoner.
"Where a distinct verbal response
cannot be achieved through apparent sleep the prisoner will on all
occasions be roused as follows:
The Rousing Procedure
Custody staff undertaking checks
of prisoners should use the following checklist where prisoners require to be
roused.
·
Can they be woken?
·
Go into the cell
·
Call their name
·
Shake them gently
·
Response to questions - can they give appropriate answers to questions
such as:
-
What is your name?
-
Where do you live?
-
Where do you think you are?
·
Response to commands - can they respond appropriately to commands such
as:
-
Open your eyes
-
Lift one arm, now the other arm.
·
Remember - take into account the possibility or presence of other
illnesses, injury or mental condition. A
person who is drowsy and smells of intoxicants may be suffering from the
following:
-
Diabetes;
-
Epilepsy;
-
head injury;
-
Drug intoxication or overdose;
-
Stroke.
Prisoners who are classed as non-vulnerable [my
emphasis] should be visited hourly to check on their wellbeing but there is no
requirement to obtain a verbal response, unless a visual check gives cause for
greater concern".
Chief Inspector Tonks
acknowledged that the words "low vulnerability" required to be substituted for
"non-vulnerable" in the last paragraph of this new section.
[55] The Chief Inspector spoke about a change that has already been
made to the Cell Sheet, restricting the requirement for a verbal response to
"vulnerable prisoners" only. He
acknowledged that this too required further amendment to reflect the new twofold
classification, otherwise it might be thought to include
[56] Finally, Chief Inspector Tonks addressed the matter of
training, and the criticism levelled at it by Sergeant Wigley. The sergeant was somewhat critical of the
training regime: in his opinion it was
"basically a straight lift from the prisoner care manual" and it was "far too
much on one day to hit people with Power Point slides". The Chief Inspector disagreed. He told the inquiry that there was a full day
of training before staff were deployed as individuals caring for prisoners,
plus another day of first aid training.
Some 45 minutes were spent on administrative matters, and the rest
of the day was focused solely on prisoner care.
There were inputs by practising custody supervisors, and by experts or
supervisors within the relevant fields.
The training was drawn from the Tayside SOP guidance, and built upon it
by utilising medical advice and guidance on how to identify vulnerable
individuals and how to care for them. It
also included information gleaned from "near misses" and fatalities.
[57] Chief Inspector Tonks acknowledged that there was no
examination, and individual knowledge was assessed by the trainers asking
questions in the course of the day's training.
However he said that training was currently being reviewed in the light
of the revised guidance, and the matter was currently in the hands of the staff
development unit. All staff in custody
teams were subject to ongoing assessment, with a formal assessment each year
where issues were highlighted. As a
consequence of that, retraining might be deemed appropriate. There was also refresher training every 3
years, but that was in the process of being reviewed.
[58] Chief Inspector Tonks did not agree that too much information
was imparted to custody care assistants in a single day. The view of the Custody User Group was that
the amount of training was appropriate - indeed it had been "bench marked"
against the training provided elsewhere.
As regards custody sergeants, the Group felt that it was appropriate for
them to have a greater level of understanding, and consequently an additional
day's training was proposed. That was
intended to be on a "one to one" basis, with the opportunity to discuss issues
with experts in the field. Apart from
that, anyone who wanted to see the Tayside SOP guidance could access it on the
Tayside Police Intranet, and a hard copy was also kept in the custody suites at
[59] I now turn to the evidence led on behalf of the deceased's
brother, Mr Frank McLellan. Some
extremely helpful comment and comparative information on practices elsewhere in
[60] Incidentally, Dr Guy's own evidence was that even if the
deceased had told him he had taken a significant quantity of methadone or
dihydrocodeine, it would probably have made no difference to the way he dealt
with him. He said that there was always
an element of "risk management" in such a situation, as one never knew whether
a drug addict might "go off" - i.e. whether he might collapse into
unconsciousness. Apart from that,
prisoners often lied about taking drugs so that they were sent to hospital
rather than locked up. Dr Guy
considered that it was the correct decision to have the deceased monitored at
police headquarters. Almost everyone he
saw in the cells had similar problems of drug-taking and slurred speech, and if
they erred on the side of caution it would mean that everyone would have to be
sent to hospital. It was rare for death
to result, and in any event, although the deceased was obviously a cause for
concern, he had been up and about in his cell when he last saw him.
[61] Dr Morrison considered that addicts could "go off"
quickly. Professor Busuttil agreed,
and added that this could happen at any time.
For him, "going off" meant difficulty with breathing, abnormality of
heart action, a rising or falling pulse rate and a reduction in the level of
consciousness to below 10 on the Glasgow Coma Scale. In his opinion, depending on the person and
the conditions at the time, the process could be "very sudden, acute and
unannounced", or a "gradual, insidious" shift from consciousness to
unconsciousness. Importantly, he
emphasised that a doctor or nurse would not be able to determine in advance how
fast deterioration might occur.
[62] Dr Morrison considered that Dr Guy's actions were appropriate
in the light of his findings. It was not
feasible to admit every drug addict to hospital, and his actions were those
expected of an experienced police surgeon.
Professor Busuttil (a police surgeon for 28 years) agreed. He considered that Dr Guy's examination was
adequate and well above average, although not entirely compliant with the
guidelines issued by the (then) Association of Forensic Physicians. (He thought that the time and ingestion of
the drugs should perhaps have been pursued further). His opinion was that quarter-hourly
observations were adequate if carried out appropriately, with a verbal response
being obtained and recorded in writing on each occasion. In the present case the deceased had appeared
fit at the time of the medical examination, and there was therefore no good
reason why he should have been sent to hospital.
[63] Dr Morrison acknowledged that no monitoring system was perfect,
even in hospital. The hope was that any
deterioration might be identified quickly enough to allow appropriate
intervention, such as a transfer to hospital, the administration of an
antidote, or resuscitation. She remarked
that some police surgeons - herself included - attached a very small "pulse
oximeter" to a prisoner's finger to indicate pulse rate and how well oxygenated
they were. Oxygen levels below 90% meant
that hospital transfer was required, whatever the level of consciousness. That said, she did not believe that an oximeter
could replace the need for a clear verbal response as an indicator of the level
of consciousness.
[64] Dr Morrison identified some other practical measures that might
well improve the care of prisoners in custody.
She referred in particular to the storage of opiate antidotes such as
Naloxone (apparently available in her area), a computerised clinical record to
enable a prisoner's previous medical records to be accessed, and resuscitation
equipment. In her view the lack of such
drugs and equipment meant that what custody staff could achieve by monitoring
was very limited. Naloxone worked quite
well with heroin, although not with the longer-acting heroin, and its effects
wore off after about 2 hours. It
acted as a "safety net" to keep the brain oxygenated until paramedics arrived -
if a prisoner turned blue, or became unconscious, or there was deep snoring
indicative of severe brain stem depression, or even if he became very drowsy
with pin-point pupils, the administration of Naloxone under the skin by a nurse
or a doctor reversed the effect of the opiate and stimulated breathing. Dr Morrison told the inquiry that
although the current view was that the drug required to be clinically
administered, the British Medical Journal had recently suggested that friends
of drug addicts should be trained to administer it. She thought that custody care assistants
could perhaps be similarly trained, although hospital admission would be
required thereafter because of the need for cardiac monitoring.
[65] Dr Morrison was of the view that a very large-scale change in
prisoner care was required to make a considerable impact on deaths in
custody. There was some hope that
"Custody Suites" with clinical staff, joint training, clear protocols about
admission and monitoring, adequate drugs, resuscitation equipment,
computerisation and the sharing of notes, might achieve that result. Custody suites of this kind existed in some
inner city areas of
[66] Dr Morrison acknowledged that custody suites raised the
question of police resources, since they were more expensive than traditional
care by doctors and custody officers.
They required a group of police surgeons on call, as well as a team of
nurses with specific training. However,
she believed that they went a considerable way towards resolving the issues of
effective monitoring and confidentiality.
She felt that if the deceased had been handled in a custody suite, and
if certain key signs had been observed - deep snoring, blue or pale lips
(indicating lack of oxygen) and cold hands (indicating peripheral shutdown) -
nursing intervention might have allowed him to survive.
[67] Professor Busuttil's main conclusion in his report was that the
monitoring of the deceased had not been as rigorous and frequent as was
intended. For him this raised questions
about the training of both civilian and uniformed custody staff. He referred in particular to the
"staff-to-detainee" ratio during busy periods, and to the appropriateness of
detaining drug addicts in custody without access to medical care on a constant
basis. Like Dr Morrison, he also
referred to the value of administering Naloxone. It was effective even if a person was in a
coma, although less so when dihydrocodeine was also present.
[68] Professor Busuttil commented on the Cell Sheet entries made by
Mr Innes at 0306 hours and 0359 hours.
In his opinion, obtaining a grunt from someone did not mean waking him
up. A person who was spoken to when in a
coma might change the rhythm of his breathing, or even grunt, but still be
unconscious. It was only if he
responded with words - what Professor Busuttil described as a verbal utterance
- that he could be said not to be comatose or pre-comatose. He thought that the deceased might well have
been unconscious, or in a coma, at 0306 hours, and it was indeed possible
that he was drifting into a coma at 0204 hours. Although it was not inevitable that he would
have died from the cocktail of drugs he had taken, it was a very strong
possibility. If it had been recognised
that he was slipping into a coma, or indeed was in one, then the administration
of intravenous Naloxone and hospitalisation would have been the correct response.
[69] Professor Busuttil also commented on Sergeant Wigley's
observation that rigor mortis had obviously set in. His opinion was that if the deceased was
indeed wakened at 0306 hours, as Mr Innes' entry indicated, rigor mortis
was unlikely, because it usually required two to three hours from the time of
death to be observable. The only reason
for rigor mortis developing so quickly - apart from a seizure - was the warmth
of the room, but it would then have to be "hothouse warm". If rigor mortis had set in by
0440 hours, then death would have occurred some two to two and a half
hours earlier - in other words, by 0240 hours at the latest. However, since training was required to
interpret rigor mortis, Professor Busuttil considered it quite likely that
the signs might simply have been misinterpreted or over-interpreted.
[70] With reference to Mr Innes' entry at what turned out to be
0359 hours, Professor Busuttil was asked whether the position of the
deceased on that occasion was a cause for concern. He observed that the deceased was in effect
in a recovery position, and consequently he would not have choked on his own
vomit (of which there was evidence after death). However, he said that if he had been the
doctor dealing with the deceased, and found him in that position, he would have
wanted to place him in an upright position, examine his responses, eyes and
chest, ascertain his level of consciousness and act accordingly.
[71] Professor Busuttil told the inquiry that this would have been
the reaction of nurses under a new system introduced in the Lothian and Borders
Police area. Following a successful,
six-week pilot in which nurses had assisted police and custody staff in the
monitoring of prisoners, money had been found from the police budget to enter
into a contract with NHS Lothian for the supply - over a five-year period - of
two highly qualified G-grade nurses between 7pm on Fridays and 7am on Sundays,
and one nurse each evening between 7pm and 7am. The nurses were based at St Leonards police
station, but could be called out to other detention centres within the Lothian
and Borders area. The annual cost of the
contract was some £1.7m.
[72] Professor Busuttil said that a scheme of this kind was not new
- it had existed for many years in
[73] Professor Busuttil went on to describe the Lothian and Borders
scheme in more detail. As I understood
him, persons with medical or mental health complaints, or with drug and alcohol
difficulties, are referred by the duty sergeant to the nurse for
assessment. The sergeant and the other
custody staff are not involved beyond that referral - it is the nurse's job to
monitor the prisoner within the cell complex.
This is done by monitoring the level of consciousness, by obtaining
verbal responses and also by checking the pupils and the pulse rate. A nurse may also communicate with the doctor
on call, and between them they may institute regular observations. These may range from quarter-hourly to
two-hourly checks, depending on the risk assessment. If a prisoner is about to slip into a coma,
a nurse is authorised to administer Naloxone.
Importantly, Professor Busuttil emphasised that the final decision
on the monitoring of a prisoner is taken by the nurse and the doctor, not by
police or custody staff. Nurses operate
autonomously, to the extent that if they consider there to be an emergency, the
prisoner concerned is taken to hospital without the involvement of custody
staff.
[74] Professor Busuttil was asked whether there had been any
problems with the initial referral by the duty sergeant. I understood from his answer that there had
not been. He said that detainees were
asked certain questions about their alcohol use, their drug use and their
medication. Any detainee who, when
observed or questioned, exhibited a potential problem with alcohol or
drugs was automatically seen by the nurse.
Professor Busuttil said that this had resulted in a decree in
admissions to Accident and Emergency departments, and it might also have saved
police time, although no cost-benefit analysis had been conducted. He also made the important point that the
police themselves looked on the new regime as a safety net - both police
officers and custody staff could rely on the presence of a qualified person to
monitor the problems that drug and alcohol users presented.
The submissions
[75] After the evidence was
concluded, the parties lodged detailed written submissions. The Crown's position, in essence, was that
there was more than enough evidence to enable a formal determination to be made
that the cause of death was methadone and dihydrocodeine toxicity, combined
with chronic alcoholism. The real
question was whether the court should make findings under section 6(1)(c) and (d)
of the Act. Dr Griffiths submitted
that since the quantity of methadone alone was potentially lethal, and its
effect was increased by dihydrocodeine, the deceased's death would appear to
have been likely, if not inevitable, and the ultimate responsibility for it lay
with the deceased himself. The police
had a duty of care to persons in their custody, but they could not be blamed
for the deceased's own actions or lifestyle.
[76] Dr Griffiths submitted that the actions of Sergeant Copeland
and Mr Penman had been impeccable.
The only trivial point was their failure to record the exact words used
when Mr McLellan was spoken to, and nothing turned on that. Dr Guy had also done as much as could
be done for an unco-operative prisoner.
Matters had only begun to go awry after Sergeant Wigley's admitted error
in writing "non-vulnerable" on the Custody Record. Although this was clearly unfortunate, and
ought not to have happened, it had never been suggested that it represented
more than a straightforward human error by a busy officer working under
considerable pressure.
[77] Dr Griffiths submitted that Sergeant Wigley's error did
not represent a defect in the system of working. Its main consequence was that on a number of
occasions the deceased had been checked visually by Mr Innes as a
non-vulnerable prisoner, without a response being required. These checks had been appropriate for a
non-vulnerable prisoner. More
importantly, there had been no significant change to Mr Innes' belief that
the deceased was still alive at 0359 hours, and no evidence other than that of
Professor Busuttil had been led to contradict it. Had he been classified as "highly
vulnerable", some form of response would have been required at each visit. The ideal response was verbal (which was the
literal requirement of the Tayside SOP) but depending on circumstances, and for
obvious humanitarian reasons, the movement of an arm or something similar would
be acceptable to front-line staff, if not necessarily to Chief Inspector Tonks.
[78] Turning to the Tayside SOP, Dr Griffiths submitted that Chief
Inspector Tonks, who was a very senior operational police officer, was quite
clear that it was adequate if followed correctly. The only change of real significance that was
proposed was the introduction of the specific "rousing procedure" for highly
vulnerable prisoners, based on English practice. However, the success of such a procedure
depended on the prisoner being properly classified as highly vulnerable in the
first place, and it could not be said that such a procedure, even if followed
to the letter, would necessarily have prevented Mr McLellan's death.
[79] Dr Griffiths considered that Dr Morrison's evidence was helpful
in highlighting the real practical difficulties faced by clinical and custody
staff - particularly the speed with which prisoners on drugs could "go off" -
but he doubted its relevance to the deceased's death. Her views had not been put to Chief
Inspector Tonks, and there was no evidence about the extent to which Tayside
Police were proposing to change their custody procedures. If the changes that she suggested were to be
implemented, there would clearly be implications for police resources in terms
of personnel and accommodation.
Dr Griffiths made similar comments about Professor Busuttil's
evidence. It was not known whether
Tayside Police had plans to follow the Lothian and Borders example, or how such
a scheme might work in their area. For
instance, what volume of custodies would justify basing a nurse in
[80] Dr Griffiths submitted that on the evidence there were no
reasonable precautions whereby the death of Mr McLellan might have been
avoided. It was not easy to see how the
possibility of isolated human error could ever be completely eliminated by
precautions that were in themselves reasonable. Joint decision-making between duty sergeants
was impracticable and unreasonable; a
continuous computer record was not foolproof against such an error; and in any event there was no clear evidence
before the inquiry that the deceased would necessarily have survived had
Sergeant Wigley not made his unfortunate error.
[81] That apart, there was no evidence to justify a finding, on the
balance of probabilities, that Sergeant Wigley's error had contributed to the
death, particularly since the checks had been carried out more frequently than
was strictly necessary for a non-vulnerable prisoner. Nor was there any evidence that the outcome would
have been different had the checks been carried out at precise half-hourly
intervals. Dr Griffiths described
Sergeant Wigley's error as an isolated, individual mistake, and submitted
that it did not represent a defect in the system of working. The deceased had twice been seen by
Dr Guy, and the checks carried out were broadly in line with his
recommendations. All that could be said
was that at some point between 0306 hours and 0440 hours - more
probably towards the end of that period - the deceased was overwhelmed by the cocktail
of drugs that he had voluntarily ingested.
In these circumstances there was no evidence to justify a finding under
either section 6(1)(c) or 6(1)(d) of the 1976 Act. Accordingly Dr Griffiths invited me to
make a determination in terms of section 6(1)(a) and 6(1)(b) only.
[82] On behalf of the Chief Constable, Mr Reid submitted that no
drugs had been taken by the deceased after his detention at 1925 hours, and a
full strip search had found nothing on him.
Sergeant Copeland had carried out a full and proper assessment of his
vulnerability in terms of the Tayside SOP, and his initial assessment of high
vulnerability and half-hourly checks had been appropriate. Dr Guy had then examined the deceased, but
as was common with someone who had taken drugs, he had been unable to obtain
specific details from him.
Nevertheless, when he examined him at approximately 2200 hours, he
had seen no need to change his previous instructions. Mr Reid noted that Dr Morrison had not
criticised Dr Guy's actions, which she considered to be of an appropriate
standard. He also pointed out
(correctly) that Professor Busuttil's report was mistaken in stating that
Dr Guy asked for quarterly hour observations of the deceased's pupils,
pulse and blood pressure. (No evidence
was led on that, and I did not take it into account).
[83] Mr Reid submitted that Sergeant Copeland had been happy
with Sergeant Wigley's decision to reduce the observation period to
half-hourly, in line with Dr Guy's instructions, This ought to have meant that he retained
his HV status, but Sergeant Wigley had of course mistakenly altered it to
NV. Although he had admitted the
mistake, he could not explain why it was made.
Mr Reid submitted that it appeared to have been a mistake made "in the
heat of the moment", perhaps because Sergeant Wigley was altering the length of
the observation period. It had a
"knock-on effect" in Mr Innes then treating the deceased as non-vulnerable,
with no specific response being required, even although the half-hourly checks
were appropriate to HV rather than NV status.
However there had been little criticism of Mr Innes, although Chief
Inspector Tonks had expressed the hope that a custody care assistant might
have noticed that NV status and half-hourly checks did not go together in terms
of the Tayside SOP.
[84] Mr Reid submitted that the Tayside SOP required a verbal
response from a HV prisoner, and Chief Inspector Tonks had said this meant
exactly what it said: an actual verbal
response had to be obtained and then entered on the Cell Sheet. However the other police officers and
custody care assistants had expressed the view that a response such as a grunt
or the movement of an arm was acceptable;
and some had mentioned human rights and the extent to which it was
appropriate to disturb a prisoner.
[85] Turning to the timing discrepancies, Mr Reid submitted that
checks had been carried out by Mr Innes at approximately half-hourly intervals
until 0306 hours, when he had wakened the deceased and obtained a response from
him. It was accepted that the next
check was carried out at 0359 hours, not at 0345 hours as the Cell Sheet
stated, but on this occasion Mr Innes had said that he could see chest movement
and had been happy that the deceased was just asleep. It had not been suggested to him that the
deceased was anything other than alive at that point. It was also accepted that the stated check
at 0420 hours was an error, and it was in fact carried out at 0440 hours. However, Mr Innes had immediately gone for
help, and there was no question of him having delayed in doing so.
[86] Mr Reid submitted that the deceased's decision to abuse his
body through consuming drugs and alcohol had effectively resulted in him
killing himself, and there was no evidence to justify any findings under
sections 6(1)(c), (d) or (e) of the 1976 Act.
There was no evidence to support the contention that if the checks had
been carried out at exactly half-hourly intervals, this would have avoided Mr
McLellan's death. Furthermore, Sergeant Wigley's
mistake had been a "one-off", and there was no evidence to suggest that there
was some reasonable precaution that could eliminate human error. Whilst it was clear that those responsible
for the care of vulnerable prisoners should be obliged to follow detailed
procedures to reduce to a minimum the chance of any death in custody, there was
no system of work that could ever completely avoid the possibility of human
error. In any event, there was no evidence
to support a finding that his mistake had caused, or contributed to, the
death. Based on the evidence, all that
could be said was that the deceased had died some time between 0306 hours and
0440 hours, although it was possible to argue that death might have occurred at
some point between 0359 hours and 0440 hours.
[87] Mr Reid submitted that the procedure in the Tayside SOP for the
obtaining of a distinct verbal response was sufficient provided it was followed
properly. The only noteworthy change
that had been proposed was to introduce for highly vulnerable prisoners the
"rousing procedure" that had been introduced in
[88] Mr Reid felt that Dr Morrison's evidence was more
appropriate for consideration in a national context. She had accepted that much of what she had
suggested depended on establishing and maintaining links between various
agencies, and in particular on appropriate funding. She had also accepted that factors such as
the geographical area covered by a police force were highly relevant. However, the various matters to which she
referred had not been raised with Chief Inspector Tonks, and there was no
evidence that any of her suggestions would have been appropriate in the context
of the present case. In particular
there was no evidence to suggest that the absence of custody suites, or the
absence of clinical staff in police stations, constituted a defect in the
system of work. Nor was there any
evidence that the existence of suites or the presence of clinical staff amounted
to reasonable precautions whereby the death might have been avoided.
[89] Mr Reid commented on Professor Busuttil's evidence that if
there were signs of rigor mortis when the deceased was found at 0440 hours,
death would have occurred by around 0240 hours at the latest. He submitted that this could not be so,
since Mr Innes had roused the deceased and obtained a response - a grunt
or something similar - at 0306 hours, and there was no suggestion in the
evidence that he had been mistaken.
Professor Busuttil had said that a grunt did not preclude the
possibility of a person being in a coma, but what it did mean was that the
person concerned was still alive.
Mr Reid considered that the inevitable conclusion was that anyone
who thought he saw signs of rigor mortis at 0440 hours was wrong. The evidence of such signs had come from
Sergeant Wigley, but not from Dr Guy, and the conclusion therefore
had to be that Sergeant Wigley was wrong.
[90] Finally, Mr Reid submitted that there were no grounds for finding
that the absence of the arrangements in operation in Lothian & Borders
reflected any defect in the system of work in Tayside, or any failure to take
reasonable precautions. Looked at
nationally, the arrangements might well be a potential way ahead for many
police forces, but factors such as cost, the numerical size of a police force,
and the size of its geographical area would be highly relevant to the
introduction of such a scheme. The
annual cost of £1.7 million over a five-year contract was a significant factor
in the Lothian and Borders budget, and there was insufficient evidence to draw
any definite conclusions on whether similar arrangements were appropriate for
Tayside Police or indeed any other force.
According to Dr Morrison, even Strathclyde, the country's largest
Police Force, had apparently not yet gone down that route.
[91] On behalf of Mr Francis McLellan, Mr Murphy submitted that the Tayside SOP required a distinct verbal response to be obtained from every HV prisoner, with the words spoken being recorded on the Custody Record. Yet between 2100 hours and 2300 hours Mr Penman had recorded nothing about what the deceased might have said. The whole purpose of the procedure was presumably to monitor a person's level of consciousness, but if no information was entered, no pattern of deterioration or improvement could be shown. The omission of such information was indicative of a breakdown in the implementation of the correct observation procedures. Mr Murphy also referred to the inconsistency between the Tayside SOP and the Cell Sheet about when a verbal response was required, with the former requiring a distinct verbal response only in the case of HV prisoners and the latter requiring a verbal response on every visit.
[92] Turning to Sergeant Wigley's error, Mr Murphy submitted that it appeared from the evidence that Sergeant Copeland was unaware of the change from HV to NV status. That had been confirmed by Sergeant Wigley, who stated that although they had both discussed the deceased at some length, no mention had been made of his vulnerability status. As there was no system in place to ensure that such an error could be avoided, the result was that the "wrong system of observation" had been used to monitor the deceased. Sergeant Wigley had also given evidence that such mistakes had been made in the past (albeit not by him). Chief Inspector Tonks had admitted that there was no requirement for an outgoing officer to verify the entries of an incoming officer, but if such a system had been in place, requiring Sergeant Copeland to initial the entries made by Sergeant Wigley, it was reasonable to anticipate that the latter's incorrect entry would have been identified.
[93] Because no such safety check was required, the result was that the deceased had been treated as "non-vulnerable" from 2300 hours until his death was discovered at approximately 0440 hours. The Tayside SOP provided (page 21) that any matter relating to a prisoner which caused concern to custody personnel should immediately be brought to the attention of a supervisory officer. It also provided (at pages 14 and 21 respectively) that any prisoner under the influence of drugs was automatically considered highly vulnerable, and any prisoner on half-hourly checks was classed a highly vulnerable. In these circumstances, Mr Innes ought to have addressed the anomaly of a NV prisoner on half-hourly checks at the earliest opportunity. His failure to do so had had a serious impact on the events that followed, and raised issues about his knowledge of the way prisoners were classified.
[94] Mr Murphy acknowledged that Mr Penman appeared not to have mentioned to Mr Innes at the handover that the deceased was under the influence of drugs. Had that been discussed, it might have alerted Mr Innes to the anomaly in the deceased's status. Mr Murphy submitted that this appeared to be a breakdown in the system of communication, as it was contrary to the instruction in the Tayside SOP (page 1) that knowledge of each prisoner was critical to their safe custody, "and information must be shared through full and accurate custody recording". Furthermore, the Tayside SOP stated (page 21) that officers in charge were to ensure that support staff were fully briefed about their specific duties in respect of each prisoner, and outgoing custody officers and custody care assistants were to ensure that the incoming assistants were fully briefed about every vulnerable prisoner. Yet it was clear that Sergeant Wigley had not specifically mentioned to Mr Innes that the deceased was to be classed as highly vulnerable. Had he done so, it might have cleared up any confusion.
[95] Mr Murphy then addressed the issue of staffing on the night of
the deceased's death.
Sergeant Wigley had remembered releasing nine prisoners in the
course of the night, and that it was Mr Innes who was responsible for
bringing them from the cells. He had
also remembered that prisoners were released at 0328 hours and 0335 hours. However, at around 0336 hours Mr Innes
ought to have made a scheduled visit to the deceased - a visit that did not
actually take place until 0359 hours.
Despite it being a busy evening, and despite Mr Innes being
involved with other prisoners, there had never been more than one custody care
assistant on duty, and this had been the norm during the two years that
Mr Innes had been a custody care assistant at
[96] Mr Murphy submitted that the circumstances of Mr McLellan's death were complex and whether or not his death could have been avoided was not to be explained by one particular oversight or system failure. It was a combination of errors and shortcomings which, when analysed together, showed that his death was not inevitable and could have been avoided. He enumerated these errors and shortcomings as follows:
(a) Sergeant Wigley's error had a crucial effect on the way he was subsequently monitored, with no distinct verbal response being required. It was clear that this error was not an isolated instance.
(b) Mr Innes' failure to identify the anomaly compounded the initial error. It indicated a lack of knowledge on his part that under the Tayside SOP prisoners on half-hourly checks or under the influence of drugs were to be deemed highly vulnerable. Had he been aware of that, he could have brought it to the attention of Sergeant Wigley, and Mr McLellan would have been monitored as a highly vulnerable prisoner. But even if he had been aware, it was clear that the Tayside SOP was not fully known or understood by custody care assistants. At no time was any response by Mr McLellan noted, albeit this was information that could play a crucial role in assessing the physical and mental health of those in custody.
(c) The level of work that Mr Innes was required to undertake might account for the slippage of time in monitoring the decease, and whilst that in itself might not have a direct bearing on whether his death could have been avoided, it continued to create an impression of a system where very little margin for error was available.
(d) The NCPE Guidance, produced on behalf of
ACPO and the Home Office, was intended to assist agencies across the
(e) There was no scheme of the kind that operated in the Lothian and Borders area. It was clear from Mr Innes' evidence about entering the cell at 0306 hours that, unlike medical staff, he was not in a position to identify any danger signals. That was of particular concern in circumstances where there was no way of telling when a patient might "go off". With the provision of proper and adequate medical care, including Naloxone, the deceased's chances of survival would have increased. Trained nursing staff would have been better able than Mr Innes to monitor his level of consciousness at 0306 hours, when a grunt was obtained and he did not physically move. A system involving medical staff would not only have acted as a safety net, but would also have relieved both police officers and custody care assistants from making decisions that they were not qualified to make.
The Determination
[97] Section 6(1) of the 1976 Act requires me to make a determination setting out the following circumstances of Mr McLellan's death, so far as they have been determined to my satisfaction: (a) where and when the death and any accident resulting in the death took place; (b) the cause or causes of such death and any accident resulting in the death; (c) the reasonable precautions, if any, whereby the death and any accident resulting in the death might have been avoided; (d) the defects, if any, in any system of working which contributed to the death or any accident resulting in the death; and (e) any other facts which are relevant to the circumstances of the death.
[98] The day and place of Mr McLellan's death are not in doubt. I was satisfied that he died in the early morning of 24 December 2005, in Cell 4 at the Divisional Police Headquarters in Barrack Street, Perth. It is less clear precisely when he died, but I am also satisfied on the evidence that it can be narrowed down to a specific period of time. Although Dr Guy pronounced him dead at 0505 hours, I conclude from the testimony of Mr Innes and Sergeant Wigley that he was almost certainly dead when seen in his cell at 0440 hours. Accordingly, the terminus ad quem for his death was 0440 hours.
[99] Secondly, I am satisfied on the basis of Sergeant Petrie's investigation that the previous visit to Mr McLellan took place at 0359 hours, and not at 0420 hours or 0345 hours as the Cell Sheet states. That was not disputed by any of the parties. For the avoidance of doubt, I do not consider there to be anything untoward about these mistaken things. I accept that more prisoners than usual were released that night; that Mr Innes was the only custody care assistant on duty; and that he had to combine both monitoring and processing functions. Mr Innes himself thought that he might have written down the wrong time because it was a busy night, and I am inclined to think that this is indeed the most likely explanation.
[100] The evidence about Mr McLellan's condition at 0359 hours came from Mr Innes alone. However, I considered him to be a good and careful witness, and I accept his evidence that at 0306 hours he roused Mr McLellan, and obtained a response - in the nature of a grunt or an "ahah" - from him. I also accept his evidence that when he visited the cell again at 0359 hours, he saw a chest movement, and was satisfied that Mr McLellan was breathing, albeit asleep. So whether or not he was in a coma, he was clearly alive at that point. Accordingly I conclude that the terminus a quo for Mr McLellan's death was no earlier than 0359 hours. He therefore died at some point between then and 0440 hours.
[101] It will be apparent from this conclusion that I reject Sergeant Wigley's evidence about rigor mortis at 0440 hours. Professor Busuttil's undisputed evidence was that this usually takes two to three hours to develop, unless a seizure has occurred or it is "hothouse warm". There is no evidence for either of these events. It is noteworthy that Dr Guy - the person best qualified to observe and record any signs of rigor mortis - made no mention of them at all. Furthermore, as Professor Busuttil observed, training is required to interpret such signs, and Sergeant Wigley's observations are necessarily those of a lay person. In these circumstances, and given my conclusion about the time of Mr McLellan's death, I am quite satisfied that rigor mortis could not have set in by the time he was discovered at 0440 hours.
[102] An accident may be defined as anything that happens without foresight or expectation (OED). Insofar as there can be said to have been an accident that resulted in Mr McLellan's death, it was his ingestion of a dangerous quantity of drugs. That this had catastrophic consequences is plain from all the medical evidence. Death was caused by a toxic cocktail of methadone and dihydrocodeine - both of which are opiates - with chronic alcoholism and consequential cardiac enlargement as a significant contributory factor. The evidence points unequivocally to such a conclusion: the toxicological analyses revealed a potentially lethal level of methadone in his body; medicine bottles (including one with methadone residue) were found on his person when he was arrested; he himself admitted taking drugs, including diazepam; and at 2100 hours he was observed to have unreactive pupils, which is a sign of opiate use. Apart from that, and although he was more sober than usual, he was unsteady on his feet, with slurred speech, and Dr Guy concluded that he had obviously been drinking. Furthermore, there was no other obvious reason for his death: there were no injuries on his body, and indeed Dr Sadler described the external post-mortem examination as "unusually negative".
[103] Both Dr Sadler and Professor Busuttil explained in detail the precise mechanism of Mr McLellan's death: essentially, the combination of opiates depressed the vital respiratory centre within the brain stem, culminating in respiratory arrest. It is clear from their evidence that a catastrophic shutdown of this kind may occur even in someone who is a habitual user, and tolerant of the drugs. Mr McLellan's history of drug use, and what was found on his person that night, points to him being such a person. So what is the risk of someone dying in such circumstances? Dr Guy commented that death is rare, and this may well be so in comparison with the numbers who abuse drugs. But Dr Morrison observed in her report that Mr McLellan was one of several people to die in police custody that year, and Professor Busuttil remarked that death, albeit not inevitable, was certainly a "strong possibility". I accept that statement as a concise characterisation of the situation.
[104] Nevertheless, although I conclude that Mr McLellan's death that evening was a strong possibility, there was nothing about his physical condition that could - or should - have given those who dealt with him any inkling about what was about to happen. He exhibited signs of drug and alcohol use, but sadly there was nothing that differentiated him from the routine drug and alcohol abuser who is a familiar feature of weekend life in police stations throughout the land. It is particularly unfortunate that he himself did little to enable his exact condition to be understood. He was vague and unco-operative about what he had taken - indeed he denied taking any drugs in the previous 24 hours, which was almost certainly untrue. His behaviour was singularly unhelpful to both Sergeant Copeland and Dr Guy, who were doing their best to assess and assist him. Had he been more forthcoming about his actual drug intake, it is just possible that his condition might have been better understood, and there could have been a more precise assessment of risk.
[105] The risk with a drug addict like Mr McLellan was spoken to by all the medical witnesses. It is that such a person may "go off" - that is, deteriorate - without warning Professor Busuttil described the clinical symptoms: difficulty with breathing, abnormality of heart action, a rising or falling pulse rate and a decline in the level of consciousness to below 10 on the Glasgow Coma scale (the norm being 15). However, a particularly important feature of his evidence, which I accept, is that the speed of deterioration varies, and it cannot be predicted, even by a doctor or nurse. As he said, death may be very sudden, acute and unannounced; or it may involve a gradual and insidious decline in the level of consciousness. It is therefore theoretically possible that death struck Mr McLellan quickly between 0359 hours and 0440 hours that weekend morning. But equally it is theoretically possible that his decline was a gradual one.
[106] The evidence does not enable us to choose between these theoretical possibilities. The simple and unfortunate truth is that we cannot determine how quickly Mr McLellan died. At 0204 hours and at 0306 hours he responded with a grunt, and at 0359 hours no response was sought. I accept Professor Busuttil's evidence that someone who grunts may still be in a comatose or pre-comatose state. Consequently, it is perfectly possible that Mr McLellan was drifting into a coma at 0306 hours - or even that his level of consciousness was declining at 0204 hours. However, in the absence of constant medical monitoring, whether or not that was so will never be known. It is just as possible that at a point between 0359 hours and 0440 hours his condition quickly and acutely deteriorated, and he died.
[107] However, what is known is that at 2300 hours a most unfortunate error was made in the assessment of Mr McLellan's vulnerability. Before then, the procedure that led to him being monitored at half-hourly and then quarter-hourly intervals, as a highly vulnerable prisoner, cannot in the main be faulted. Apart from their understanding of the requirement to obtain a verbal response (on which I comment later), Sergeant Copeland and Mr Penman behaved impeccably. I also consider that Dr Guy's examination and treatment of Mr McLellan was beyond criticism. Nevertheless, it is important to note that when Dr Guy instructed quarter-hourly checks for a couple of hours, followed by half-hourly checks thereafter, he himself appears to have said nothing about Mr McLellan's vulnerability. That was not through any omission on his part; it was because it is the duty of the supervising police officer, not of the police surgeon, to assess a prisoner's vulnerability. Yet the assessment of vulnerability is a particularly critical task, for it determines not only the frequency but also the nature of the monitoring that takes place thereafter. Whether it is a task that ought to be performed by a lay person, even if that person is an experienced police officer, is a matter to which I will return.
[108] Sergeant Wigley admitted his error with commendable frankness, and he was undoubtedly correct to do so. The Tayside Standard Operating Procedure ("the Tayside SOP") in force at the time refers (page 21) to the frequency of visits in the high vulnerability category ranging from constant observation to observation at intervals of up to 30 minutes. Consequently, Dr Guy's instruction to move from quarter-hourly to half-hourly visits after two hours should not have resulted in any alteration to Mr McLellan's HV status. There is no doubt that Sergeant Copeland acted correctly - and commendably - in ascribing that status to Mr McLellan, notwithstanding the fact that he denied using drugs or consuming alcohol in the previous 24 hours, for the succinct reason that he gave to the inquiry. But there is also no doubt that Mr McLellan's HV status should have been retained.
[109] Sergeant Wigley himself was unable to give any explanation for his error, which is unfortunate. However, I am satisfied that it should at least be seen in the context of the discussions between him and Sergeant Copeland at the handover. At that time, Sergeant Copeland was of the opinion that Mr McLellan's vulnerability was greatly reduced, and hourly checks were sufficient. Although he did not say so expressly, I understood from his evidence that he conveyed this opinion to Sergeant Wigley in the course of their discussions. If that is a correct inference, and even although Sergeant Wigley himself did not mention it, that may well have been a significant contributory factor in causing him to ascribe NV status to Mr McLellan - not least because he would have had no prior knowledge of Mr McLellan's circumstances that evening beyond what he would have obtained from Sergeant Copeland.
[110] Plainly if the error had not occurred, and Mr McLellan had retained his HV status, he would have been monitored in terms of the Tayside SOP. A distinct verbal response would have been required on each visit, with his actual words being recorded on the Cell Sheet. However even if such a monitoring regime had been in operation, and even if it had been rigorously adhered to, I am unable to conclude that it would have saved Mr McLellan's life. Death - or irreversible deterioration - could still have occurred suddenly and acutely, without prior warning, between the half-hourly visits, even if the monitoring had been done by a nurse. As Dr Morrison acknowledged, no monitoring scheme can be perfect, even in a hospital environment. With hindsight it would have been better to have left Mr McLellan on quarter-hourly observations, as that might have increased his chances of survival. With hindsight also, the most effective monitoring would have been in a hospital environment. However, Dr Guy's judgement that half-hourly observations were sufficient after two hours was never questioned, and there is no reason whatsoever to doubt its propriety. There was nothing in Mr McLellan's presentation to suggest that he ought to have been transferred to hospital, far less subjected to constant observation. As Dr Guy commented, if Mr McLellan ought to have been taken to hospital, all drug users in police stations would require to be taken there.
[111] For the purposes of this Inquiry three key questions require consideration: (1) whether there were any reasonable precautions whereby his death and any accident resulting in his death might have been avoided; (2) whether there were any defects in the system of working which contributed to his death or any accident resulting in his death; and (3) whether there are any other facts which are relevant to the circumstances of his death. I will deal first of all with the matter of reasonable precautions. I am satisfied that there were no reasonable precautions whereby the accident - i.e. the deterioration caused by the combination of drugs and alcohol - might have been avoided. Only Mr McLellan could have prevented that, by reducing his drug and alcohol intake. However, I have concluded on reflection that a reasonable precaution which might have avoided his death would have been to require an incoming duty officer to cross-check and confirm the level of vulnerability with the outgoing officer at the time of the handover.
[112] The period of a handover is particularly critical. Unlike the situation during a shift, when a duty officer alters the assessment of vulnerability by reference to his own knowledge and experience, an incoming officer at the start of a shift has to rely on information imparted by the outgoing officer. That information may be imperfectly received. On a weekend nightshift, with a number of prisoners being brought in, and possible pressures on time, the risk of that occurring is heightened. Given the sheer volume of prisoners who have to be processed annually, and the scarcity of resources, I entirely accept that it is not a viable proposition to require two supervising officers to be present throughout each shift, far less to require a third officer to make himself available as an arbiter. Nevertheless Chief Inspector Tonks did acknowledge that it is feasible for a cross-check to take place during each of the thrice-daily handovers, when the incoming and outgoing officers are together for about 15 minutes, discussing each of the prisoners.
[113] I consider that it would have been a relatively simple task for officers to discuss the vulnerability status of a prisoner, and for the outgoing officer to countersign the status before going off shift. If any change had been proposed, and it had been agreed, both officers could have signed the alteration to evidence that agreement. And if they had been unable to agree, a fairly straightforward solution would have been to obtain medical advice before taking a final decision - or, if no such advice had been readily available, to err on the side of caution in the meantime, by retaining or ascribing HV status. Had such a requirement been in place on the night of Mr McLellan's death, the whole tenor of Sergeant Copeland's evidence suggests that he would have told Sergeant Wigley to retain his HV status. Furthermore, in the light of Sergeant Wigley's acknowledgement that NV status and half-hourly observations are incompatible, it is very likely that he would have taken Sergeant Copeland's advice.
[114] It is of course impossible to conclude that this would have avoided Mr McLellan's death, for the reasons that I have already given. However - although I accept that it is only in hindsight - a procedure expressly requiring both officers to discuss and confirm the vulnerability status of each prisoner might have avoided his death. The retention of HV status would have required a distinct verbal response - in the form of actual words - to be obtained. If his decline was gradual, his inability to provide such a response might have alerted staff to the gravity of his condition. And even if his deterioration was sudden, there might have been a fortuitous intervention at a time when effective intervention - such as the administration of Naloxone or emergency paramedical assistance - was still possible.
[115] I consider that another reasonable precaution would have been to prioritise the monitoring of prisoners. I was perturbed to note that in the period between 0306 hours and 0440 hours - a period of just over an hour and a half - Mr McLellan was visited only once. That is clearly unacceptable when a police surgeon has instructed half-hourly observations, and no one has suggested otherwise. Mr Innes obtained a response from him at 0204 hours and at 0306 hours, and had there been visits at around 0330 hours and 0430 hours, in addition to the visit at 0359 hours, he might well have sought a response of some kind during at least one of these visits - notwithstanding his belief that Mr McLellan had NV status. If Mr McLellan had "gone off" at that point - whether suddenly or gradually - then the absence of any response at all might have alerted the staff to his condition, and saved his life.
[116] The next question is whether there were any defects in the system of working which contributed to Mr McLellan's death, or any accident resulting in his death. In my opinion there were none. The very human error made by Sergeant Wigley was not a defect in the system for processing prisoners set out in the SOP; it was, rather, a departure from what that system effectively required, which was the ascription of HV status to a prisoner on half-hourly observations. Nor was the failure of Mr Innes to monitor Mr McLellan timeously a defect in that system: the problem was again human rather than systemic.
[117] It can be argued that the absence of a requirement in the Tayside SOP to cross-check, or to prioritise the monitoring of prisoners, amounted to defects in the system. To the extent that a reasonable and informed observer, on perusal of the Tayside SOP, might have expected to see such safeguards in place, given the central importance of the vulnerability assessment for the well-being prisoners, I consider that to be correct. However, even if that is so, I am satisfied that these are not defects which are relevant to section 6(1)(d) of the 1976 Act. For the purposes of that subsection the question is whether there were any defects in the Tayside SOP which "contributed" to Mr McLellan's death, or any accident resulting in his death. Although it is possible to conclude on the available evidence that the existence of either requirement might have avoided his death, it is quite impossible to conclude that the absence of either requirement was a contributory factor in his death.
[118] For the avoidance of doubt, I do not consider that the absence from the Tayside system of the medical monitoring which I am told now operates in the area of the Lothian & Borders Police Force amounts to a defect in that system. The apparent substitution of medical for police monitoring in Lothian & Borders is a fundamental change. It not only involves the creation of a qualitatively different regime; it also raises questions of policy about the application of financial and other resources, and therefore about spending priorities. It is not the function of a court to determine what the priorities for public spending should be. I am satisfied that the Tayside system makes reasonable and adequate provision for the involvement of qualified medical personnel, and indeed this is well illustrated by Dr Guy's involvement in the examination and assessment of Mr McLellan. This does not mean that a system of assessing prisoners which has medical rather than lay monitoring at its heart is not likely to be more effective in protecting the well-being of prisoners. Nor does it mean that the introduction of such a system should not be an aim for the future. All it does mean is that a system which makes reasonable and adequate provision for the well-being of prisoners should not be found defective, in terms of section 6(2)(d) of the 1976 Act, because it can be criticised in the light of evolving perceptions of what such provision should entail. If that was so, every system might necessarily have to be considered inadequate and defective.
[119] These remarks form a useful bridge to the next, and final,
question for consideration, which is whether there are any other facts which
are relevant to the circumstances of Mr McLellan's death. In my opinion there are. Let me turn first of all to the procedures
contained in the NCPE Guidance produced on behalf of ACPO and the Home
Office. This was routinely described in
the course of this Inquiry as English guidance, and certainly there appears to
be no reference to any involvement by the Association of Chief Police Officers
in Scotland (ACPOS). However, as
Mr Murphy correctly observed, the preamble to the NPCE Guidance states
that it contains "procedures and practical working guides to assist agencies
across
[120] That takes me to the second issue, which is the involvement of medical personnel - doctors, nurse and paramedics - in the assessment and monitoring of prisoners. The NCPE Guidance does not insist on the primacy of any one approach. It states that there can be no "one size fits all" model for staffing levels or resource composition, although clinical healthcare provision should be available when required (page 98). It continues: "Forces should develop a healthcare model that best suits their requirements and enables them to deliver effective healthcare provision" (page 98). The various models suggested include GPs on call or employed by external suppliers: employing nurses through agencies or on call out from the NHS; paramedics on call; and protocols with local healthcare facilities (page 99). The NCPE Guidance comments that in determining the type of provision, forces should ensure the allocation of resources to enable healthcare professionals to do their job efficiently, and it then continues (page 99):
"The
presence of healthcare professionals in custody suites increases the chances of
identifying detainees who may be at risk and improves the coordination of care
for vulnerable persons".
[121] That is the heart of the matter. For me, the overarching question that lingers at the end of this inquiry is what the nature of the monitoring process should be for prisoners who abuse drugs, or alcohol, or both. Any system for the care and well-being of prisoners must recognise that we live in socially unsettling times, when alcohol and drugs have become cheaper, and their abuse more prevalent. In these circumstances there is a reasonable expectation that an increasing number of persons in custody will exhibit the symptoms of multiple drug and alcohol abuse. Accordingly, the question of healthcare provision must be at the heart of any continuing review of the procedures for the care of persons in custody. It is difficult not to conclude as a matter of theory that that the process should involve an initial assessment of all prisoners by persons with sufficient knowledge of the relevant clinical signs and patient responses to detect the vulnerable prisoner, and sufficient training to take appropriate steps to avoid or reverse any potential decline - whether sudden or gradual. I have sufficient information about the precise situation in Lothian & Borders, or indeed elsewhere in the country, to know whether there is a particular healthcare model that is favoured, or thought to be particularly effective. However, it goes without saying that it must involve trained nurses and (when appropriate) doctors, or at least trained paramedics.
[122] The proposed changes to the Tayside SOP are commendable. They place greater emphasis on the risks posed by drugs, highlight the fact that all prisoners may be vulnerable, and improve the system for monitoring them. However they do not tackle the question of how to involve healthcare personnel at the heart of the assessment and monitoring process. In my opinion - which I acknowledge is based primarily upon the issues raised in the present inquiry - the Custody User Group should make it a priority to consider whether a system of medical assessment and monitoring, perhaps along the lines of the one introduced in Lothian & Borders - and using the experience that must already have been obtained there - can be introduced in the Tayside area. As attitudes change, I suspect that the time is fast approaching when it will no longer be considered adequate for a prisoner who is at risk through alcohol, or drugs, or both, to be assessed by anyone than a healthcare professional - or at least a trained paramedic. Nor, I suspect, will it be considered satisfactory for the nature and frequency of any monitoring to be determined by a police officer, or for the actual monitoring itself to be carried out by a custody care assistant.
[123] I accept that this raises difficult questions about how such a service can be delivered in an area such as Tayside. I also accept that these questions have resource implications. However, as the NCPE Guidance recognises, the involvement of persons with healthcare experience increases the chances of identifying people who are at risk. Apart from that, and as Professor Busuttil remarked, it also allows police officers and custody care staff to lay down the heavy burden of having to carry out the assessment and monitoring of potentially vulnerable prisoners themselves, without the advantage of medical knowledge and experience. In my view this is reason enough to address in detail how much a system could operate in Tayside.
[124] That leads on to my third point.
Although the information made available to me in the course of the
inquiry was necessarily scanty, it appears that the approach to prisoner care
in
[125] A Scotland-wide system for the care of vulnerable prisoners should aim to achieve the same high standard throughout all the police force areas. This does not mean that the mode of delivery of that care would be the same in each area - obviously it is one matter to provide medical assessment and monitoring in the middle of a large city, and quite another matter to provide it in a rural area many miles from any large centre of population. However, regardless of the mode of delivery, the aim in every case should be the same: the accurate and knowledgeable assessment of risk, monitoring that is sensitive to any clinical signs of decline, and treatment that is both prompt and effective. This would serve to ensure that prisoners (and not least their families) can expect the same standard of treatment wherever they are. Importantly, it would also ensure that police and medical personnel who move from one part of the country to another will find the same system that they are already used to.
[126] I suspect that the strategic and operational resources to achieve a unified system are currently available, and simply require adequate co-ordination. There is an obvious body to achieve that. ACPOS is a collective organisation of Chief Constables, senior officers and support staff from the eight police forces in Scotland that has evolved from a staff association into a strategic body overseeing and co-ordinating all aspects of the direction and development of the Scottish police service. It could, and probably should, be the vehicle for a root and branch re-examination of the current practices and planned changes within the eight police forces, with a view to creating a system that meets the needs of a changing prisoner population.
[127] Whether or not there is to be an examination of an effective healthcare model for prisoner care, at Tayside or at national level, there are clearly measures that can be introduced now to improve the effectiveness of the Tayside procedures. They are as follows:
(a) A procedure for confirming and countersigning the vulnerability assessment of a prisoner at each handover should be introduced, for the reasons to which I have already referred. This would involve very little change in present procedures, but it should ensure that the outgoing and incoming officers focus clearly on the issue of vulnerability, particularly in circumstances where the frequency of the monitoring regime is to be altered.
(b) Priority must be given to the timeous monitoring of vulnerable prisoners, particularly those who are on quarter-hourly or half-hourly observations. Duty officers and custody officers should be made aware that processing other prisoners must not take precedence over visiting a vulnerable prisoner in his cell, unless there are exceptional circumstances.
(c) Staffing levels should be sufficiently adequate to ensure that effective monitoring can take place without undue delay in the processing of other prisoners, or undue pressure being placed on custody staff. I was concerned to note that although there was apparently supposed to be an additional custody care assistant on duty at weekends in Perth because of the greater volume of drinking, and prisoners being kept in over the weekend, that does not appear to have happened, at least on this particular weekend. No satisfactory explanation was provided for the discrepancy between Chief Inspector Tonks' understanding of the number of staff who ought to have been on duty, and the evidence of Mr Innes, who was clearly on his own. I suspect that Tayside Police may already have reviewed their staffing levels in the light of the present case, but if they have not, I recommend that they do so now.
(d) If they have not already done so, the Custody User Group might usefully focus on the question of what additional drugs, medical equipment and training can be made available to police officers and custody staff, so that they can provide as high a standard of prisoner care as is possible within the inevitable resource constraints. Dr Morrison's suggestion that lay persons - including custody care staff - might be trained to administer a potentially life-saving drug such as Naloxone in an emergency should undoubtedly be followed up. So should her suggestion that resuscitation equipment should be available, with training on how to use it. On listening to her, I also wondered whether it might be possible to train custody staff in the use of a pulse oximeter, which - albeit not fool-proof - could give an indication of low oxygen levels and thereby assist in the assessment of risk. I do not know whether these are genuinely viable propositions. Nor do I know what arrangements, if any, can be made for the supply and payment of such drugs or equipment. However, it is only when there has been a thorough assessment of whether custody care staff might be able to operate, in some respects, as paramedics that it is reasonable to exclude the possibility.
[128] Finally, there is the important matter of staff training. The evidence that I heard was somewhat disturbing, for it became plain as the Inquiry proceeded that neither the vulnerability assessment nor the procedure for obtaining a distinct verbal response was fully understood. In particular, I was not satisfied that police officers or custody staff sufficiently understood that actual words are required when seeking a "distinct verbal response" from every HV prisoner. I suspect that the lack of understanding is due to two factors in particular: (a) the somewhat ambiguous nature of the passages in the Tayside SOP on vulnerability and supervisory arrangements, and (b) the lack of clear and simple instructions at the "grass roots" level.
[129] I have to say that on reading the relevant passages in the Tayside SOP I found myself asking questions about what was meant, and could not immediately find an answer. Although the document should be in plain language, I fear that it is not. My reaction to it was that some work could usefully be done on tightening it up, and giving it greater clarity. That is a task which should perhaps be undertaken by someone with professional editorial skills who understands the requirement for plain English in official documents. If that is done, I suggest that a copy of it could be issued to police officers and custody care assistants at the end of their training. However, it is a detailed document and staff cannot be expected to retain all the information that it contains. For that reason, I suggest that I could be accompanied by a simple aide-memoire - perhaps in the form of a leaflet - containing the basic rules about the assessment and monitoring of vulnerable prisoners - such as the rousing procedure - which they could keep with them at all times, and which could also be freely available within each police station. This may seem a simple change, but in an emergency - particularly one involving inexperienced staff - it might prove to be a lifesaver. If these deficiencies can be addressed, then I do not think that the present and proposed level of training for custody staff and supervising officers described by Chief Inspector Tonks can be criticised - although staff will obviously need additional, specialised training if they are to undertake any paramedical functions.
[130] Lastly, but most importantly, it remains for me to express my very deepest sympathy to Mr McLellan's family for the tragic loss that they have suffered. Nothing can ever replace a loved one, and I appreciate how difficult and painful it must be to have the circumstances and difficulties of a person's life exposed to public view, as must necessarily be the case in an inquiry such as the present. However, I do hope that they will take some comfort from the fact that this examination of the circumstances of his death ma help to avert a similar tragedy for another family in the future.