Healthcare provision in prisons: emergency care response

This investigation focuses on the delivery of emergency care to patients in prison. Specifically, it looks at access to 999 emergency services and the ability of ambulance services to respond to 999 calls. Emergency care delivery in prisons is complicated by the environment and security restrictions that are in place. Delays in providing emergency treatment can affect the health outcomes for patients.

The investigation explored the processes in place for responding to medical emergencies in prison and how these impacted on patient safety.

This is the first of a series of reports on the theme of healthcare provision in prison; further reports will explore continuity of care, data sharing and IT, and common themes found across all areas of healthcare provision.

These investigations were launched after discussions with 26 national organisations across the healthcare and justice systems, gaining knowledge of their concerns about healthcare in prisons.

Through a patient engagement group, HSSIB engaged with over 100 patients and people being detained to understand their experiences of receiving emergency care in prison. The investigation also engaged with stakeholders from across the prison healthcare system to gather evidence about emergency care response protocols and processes from their perspectives.

• Ambulance services spent significant time diverting resources to callouts in prisons that were then cancelled, or attending medical emergencies that were not serious enough to have warranted the presence of an ambulance crew.
• Prisons are making large numbers of 999 calls for non-emergency incidents, because of a low-risk approach caused by fear among prison staff of having to attend an HM Coroner’s court and being blamed for making a wrong decision.
• No situational information about patients experiencing a medical emergency is provided direct from the scene to the 999 call handlers. Information is passed from the scene via multiple handovers before it is received by the call handlers, which can result in misrepresentation of the situation.
• Response categories of ambulances attending prisons are regularly assessed using minimal information and ambulance services spoken to therefore defaulted to category 2 (18-minute response time). This is often not the appropriate categorisation for the nature of the situation, which has delayed appropriate care to patients both in the community and in prisons.
• The emergency response card (code blue/code red card) that prison staff are given is not designed to best support staff in identifying a medical emergency and supplying the situational information that the emergency services need to triage the situation properly.
• There is no embedded recurring training to support prison staff to recognise medical emergencies that require a 999 response, to help reduce the number of calls for scenarios that are not emergencies.
• The ambulance services and the prison service have no formal communication channels with each other to enable them to review incidents and discuss policy, to ensure that they work collaboratively and efficiently together.
• Local policies on emergency vehicle access varied between the ambulance services and prisons, causing uncertainty for ambulance crews about what was required of them, from a security perspective, when arriving at a prison.

HSSIB makes the following safety recommendations

1.1.1 This investigation focuses on the delivery of emergency care to patients in prison, specifically access to 999 emergency services and the ability of ambulance services to respond to those calls. Emergency care delivery in prisons is complicated by the environment and the security restrictions that are in place.

1.2.1 The Ministry of Justice (MoJ) is the government department responsible for the justice system.

1.2.2 The justice system includes:

• courts
• prisons
• probation services
• attendance centres (Ministry of Justice, n.d.a).

1.2.3 There are 122 prisons and 500 courts within the justice system in England and Wales (Ministry of Justice, n.d.a).

1.3.1 HM Prison and Probation Service (HMPPS) is an executive agency sponsored by the MoJ. It works with partner organisations to enable the sentences of the courts to be carried out, either in custody or the community.

1.3.2 Within England and Wales, HMPPS is responsible for:

• running prison and probation services
• rehabilitation services for ex-offenders leaving prison
• making sure support is available to stop people re-offending
• managing contracts for private sector prisons and services such as:

– the Prisoner Escort and Custody Service

– electronic tagging.

1.3.3 Through HM Prison Service it manages public sector prisons and the contracts for private prisons in England and Wales.

1.3.4 Through the Probation Service it oversees probation delivery in England and Wales including through community rehabilitation companies.

Prison categories

1.3.5 Prisons are categorised according to the risks associated with the prisoners that they hold. There are four categories of prison for male prisoners:

• ‘Category A
  These are high-security prisons. They house male prisoners who, if they were to escape, pose the most threat to the public, the police or national security.
• Category B
  These prisons are either local or training prisons. Local prisons house prisoners that are taken directly from court in the local area (sentenced or on remand), and training prisons hold long-term and high-security prisoners.
• Category C
  These prisons are training and resettlement prisons; most prisoners are located in a category C. They provide prisoners with the opportunity to develop their own skills so they can find work and resettle back into the community on release.
• Category D – open prisons
  These prisons have minimal security and allow eligible prisoners to spend most of their day away from the prison on licence to carry out work, education or for other resettlement purposes.’ (Ministry of Justice, n.d.b)

1.3.6 There are two categories of prison for female prisoners: open or closed. High-risk female prisoners are classed as ‘restricted status’ and are housed in closed prisons (Ministry of Justice, n.d.b).

1.3.7 The investigation did not consider category A prisoners within the scope of the investigation due to the security restrictions placed upon them. Category A prisons were still considered in respect of the processes and procedures for the category B prisoners that they hold.

Equivalence

1.4.1 The House of Commons Health and Social Care Committee report on prison health states that:

‘Prison health and care services should be delivering standards of care, and health outcomes, for prisoners that are at least equivalent to that of the general population.’ (House of Commons Health and Social Care Committee, 2018)

1.4.2 The report recommended that the National Prison Healthcare Board (NPHB) defined what was meant by ‘equivalent’ care. The NPHB’s response and definition reads:

‘‘Equivalence’ is the principle which informs the decisions of the National Prison Healthcare Board so that member agencies’ statutory and strategic objectives and responsibilities to arrange services are met, with the aim of ensuring that people detained in prisons in England are afforded provision of and access to appropriate services or treatment (based on assessed population need and in line with current national or evidence-based guidelines) and that this is considered to be at least consistent in range and quality (availability, accessibility and acceptability) with that available to the wider community, in order to achieve equitable health outcomes and to reduce health inequalities between people in prison and in the wider community. (National Prison Healthcare Board, 2019)

1.4.3 The requirement for the equivalence statement is based, in part, on a prisoner’s inability to take themselves to an Emergency Department or to call 999 or 111 for themselves.

Prison healthcare departments

1.4.4 Each prison has a healthcare department, which is commissioned by NHS England, that provides medical services to patients within the prison. The healthcare departments provide numerous services including:

• GPs
• opticians
• dental services
• substance misuse services.

1.4.5 Healthcare departments provide different levels of cover depending on the commissioning requirements of the prison and the population. Most departments are nurse-led with GP-run clinics. Some departments provide 24/7 care, while others only provide daytime cover.

1.4.6 All healthcare departments have a role known as the emergency response nurse. The purpose of the role is to respond to medical emergencies, then assess and treat patients, while ascertaining whether further treatment and transfer to hospital is required.

1.5.1 The investigation has used the term ‘medical emergency’ to describe an incident which requires a 999 call and dispatch of an ambulance to a patient in a prison, as this was the terminology used by prison staff and HMPPS.

1.5.2 When there is a medical emergency in a prison, the immediate action is for the prison officer first at the scene to use their radio to inform colleagues of the situation.

1.5.3 HMPPS has two colour codes for emergencies:

• code red for blood/burns (for example trauma emergencies, injuries caused by incidents or accidents)
• code blue for breathing/collapses (such as medical emergencies caused by diseases or health conditions).

The ‘code red/blue’ calls were implemented following learning from Prisons and Probation Ombudsman investigations into deaths where a delay in calling 999 was identified (HM Prison and Probation Service, 2013).

1.6.1 There are 10 NHS ambulance trusts in England providing life-saving patient care and support, which includes the 999 emergency call service. They also provide other services such as non-urgent patient transport and some NHS 111 (telephone or online assessment and triage) services.

1.6.2 The ambulance trusts cover large geographical areas and are all responsible for providing emergency medical response to multiple prisons across their area.

Ambulance services triage systems

1.6.3 NHS Pathways is a clinical triage system – that is, a digital system used to help assess patients’ symptoms and direct them to the right care. It is used across all NHS 111 and some 999 services across England. Around 14 million calls a year are triaged through this system. It contains approximately 800 symptom pathways and uses standardised, universal question sets to identify the appropriate services.

1.6.4 While all NHS 111 services use NHS Pathways, some 999 emergency operations centres use a different system for triage called the Medical Priority Dispatch System.

Ambulance dispatch categorisation

1.6.5 Ambulances are dispatched under a categorisation system, with each category having a desired response time. There are 4 categories of ambulance response:

• ‘Category 1 – A time critical life threatening event requiring immediate intervention or resuscitation. Average response time target of 7 minutes.
• Category 2 – Potentially serious conditions that may require rapid assessment and urgent on-scene intervention and/or urgent transport. Average response time target of 18 minutes.
• Category 3 – An urgent problem (not immediately life threatening) that needs treatment to relieve suffering and transport or assessment and management at the scene with referral where needed within a clinically appropriate timeframe. Target of 90% to be reached within 2 hours.
• Category 4 – Problems that are less urgent but require assessment and possibly transport within a clinically appropriate timeframe. Target of 90% to be reached within 3 hours.’ (NHS Providers, 2019)

Code red and code blue

3.1.1 To prevent delays in emergency services reaching patients, national guidance was changed to reflect findings from Prisons and Probation Ombudsman investigations (see 1.5). The HM Prison and Probation Service (HMPPS) Prison Service Instruction (PSI) 03/2013, entitled ‘Medical emergency response codes’, states:

‘The intention is to ensure timely, appropriate, and effective response to medical emergencies and thereby to maximise the likelihood of a positive outcome for the patient.’ (HM Prison and Probation Service, 2013)

3.1.2 The instruction goes on to state that prisons must have a local protocol which:

‘Ensures there are no delays in calling, directing or discharging ambulances.’ (HM Prison and Probation Service, 2013)

3.1.3 This has resulted in a standardised approach across prisons where, upon receiving a code red/blue call, the control room immediately calls 999 and requests an ambulance to attend the prison.

3.1.4 Each prison officer is supplied with a medical emergency response card. This details the nature of the code blue/code red categories and what to do in the case of a medical emergency (see figure 1).

Figure 1 Medical emergency response card

3.1.5 Prison policy guided prison officers to declare a medical emergency as soon as there was any doubt about the seriousness of the patient’s condition. The prisons did not describe using the NHS 111 service as they generally had a healthcare department. Prisons’ policies did not cover the use of the NHS 111 system as they were specifically written to deal with emergencies requiring an ambulance.

3.1.6 On the back of the medical emergency response card there are instructions for prison staff: ‘What to do in a medical emergency’. The first instruction is to ‘raise the alarm by using the internal prison CODE BLUE or CODE RED call signs over the radio’. Once they have done this they are then guided to answer questions about the patient’s condition; however, by this time the 999 call has already been made without this information (see 3.2.8). The investigation also noted that prison staff were commonly placing the cards in the same lanyard card holder as their prison identification cards, which meant that they only saw one side of the card – typically the back of the card that does not have the patient condition questions.

3.1.7 The investigation conducted observations in 13 control rooms and saw the medical emergency response card pinned on boards in each one, although in every case they were pinned with only the back of the card on display. HMPPS informed the investigation that they had issued posters to prisons for control rooms to display, which had the medical questions on them. The investigation did not see the HMPPS posters on display in any control rooms, nor were they mentioned by control room staff. The control room staff did not describe asking staff at the scene to provide the information on the front of the card; not having the front of the card displayed meant there was no prompt for them to ask these questions.

3.1.8 While the card lists conditions that fit into the two codes, there is no detail about what meets the criteria. For example, ‘Severe loss of blood’ is the first category for code red, but there is no description of what severe loss of blood means, although the amount of detail able to be included is limited by the size of the card. The investigation was told, when reacting to self-harm incidents, there is a tendency that 'as soon as staff see blood they will call a code red', when the majority of cuts do not require a 999 call to be made. In addition, the two codes cover a wide variety of medical emergency situations, particularly the code blue element. This makes it harder for the staff to identify situations that actually meet the criteria for a 999 call.

3.1.9 The medical emergency response card takes a lot of information from PSI 03/2013. However, the card does not emphasise the importance of the following extract from the PSI:

‘… the member of staff using the medical emergency code must also provide relevant information about the condition of the prisoner to the control room staff, to enable them to share this with the ambulance service for use in the triage process.’ (HM Prison and Probation Service, 2013)

3.1.10 The design of the medical emergency response card does not assist prison staff in the best manner to ensure that sufficient situational information is passed to ambulance service call handlers.

HSSIB makes the following safety recommendation

Medical emergency identification

3.1.11 The prisons all employed a system where the emergency was called by the prison officers, the control room telephoned 999 and then they waited for an emergency response nurse from the prison healthcare department to go to the scene. Once the emergency response nurse had assessed the patient they would decide whether the 999 call should be stood down or if an ambulance was needed. Prison officers were not allowed to cancel the 999 call once it had been made.

3.1.12 Prison staff told the investigation that most 999 calls were eventually cancelled by the emergency response nurse. The investigation was told that the volume of calls being made by some prisons was reaching levels that hindered both the prison and the emergency services. The impact of this on the emergency services is explained in 3.2.

3.1.13 The investigation was told that a number of factors added to the volume of calls being made, including:

• A very low threshold among prison staff of what a serious injury was. Staff feared HM Coroner’s court and being blamed if a patient died, and were not prepared to make a personal judgement on whether an ambulance was required or not.
• Prison policies encouraging a low-risk approach which left staff with no room to make decisions. For example, PSI 03/2013 states that ‘it is better to act with caution and request an ambulance that can be cancelled if it is later assessed as not required’ (HM Prison and Probation Service, 2013).
• High prison staff turnover, which resulted in a lack of experience among staff who were first on the scene of medical emergencies.
• Limited staff training and knowledge of medical emergencies.

3.1.14 The investigation was told by an ambulance service that they had previously conducted some training with prison officers, giving them examples and knowledge to be able to make a more informed decision about whether an incident required a 999 call. The training had “made a difference for a while”; however, after a year the staff had either moved on, left the service or were new and the volume of calls had increased again. The improvement was not maintained as the training was only conducted once and not embedded into a recurrent training plan.

3.1.15 Some prisons had started to employ paramedics within their healthcare departments to combat the “inappropriate” code red/blue calls that were being made. There were examples where this had made a positive difference but there were also examples where it had made no difference. Paramedics told the investigation that they were mainly being used to support non urgent healthcare functions, which is not what they were trained for, didn’t make the most of their skillset, and hadn’t really had the impact they were hoping for. They had not been tasked to deliver training to prison officers, which is where they felt the most value may be added.

3.1.16 An indicator that shows the impact of the code red/blue system on ambulance services is the number of times 999 calls are made and the patient is not taken to the emergency department (ED). The investigation analysed data from ambulance trusts about how many 999 calls from prisons resulted in patients being taken to hospital, and national figures from a report by the Association of Ambulance Chief Executives (2024), which reflect all 999 calls made across England including prisons (see figure 2).

Figure 2 999 call centre data

3.1.17 The data shows that approximately 3 in 4 (71%) 999 calls were assessed as not requiring an ambulance to be dispatched to the prison. In these cases the calls were dealt with either through advice to the prison about what to do or the call being cancelled. However, for all calls made across the country, which includes the data for prisons, it was 1 in 8 (12%). Once an ambulance reached the patient only 1 in 5 (19%) needed to be taken to an emergency department, while nationally 1 in 2 (51%) patients seen by ambulance crews needed to be taken to hospital.

3.1.18 The data shows that a large number of the 999 calls being made by prisons were not for medical emergencies that required an ambulance. The figures reflect the frustrations of ambulance crews and emergency centre call handlers and dispatchers that the investigation spoke with, who stated that a significant amount of time was being wasted.

3.1.19 Some prison officers told the investigation that they were given first aid training but when asked about the emergency scenarios they mentioned the card. The training described to the investigation did not give the officers sufficient knowledge to be able to differentiate between scenarios that could be dealt with by the prison healthcare department, or were not serious enough to warrant calling an ambulance, and those that required a 999 call.

3.1.20 Since 2016 all new prison officers have completed a level 3 qualification in emergency first aid at work. While this gives new officers the ability to deliver emergency first aid, it does not cover the recognition of a life-threatening emergency which requires an ambulance.

HSSIB makes the following safety recommendation

Internal communication

3.2.1 The investigation learned and observed that internal communications in prisons were via closed network radios. All prison officers and all emergency response nurses within the healthcare team had a radio. This allowed staff across the prison, including emergency response nurses, control room staff and gate staff, to be made aware of a code red or code blue emergency through a single radio call.

3.2.2 Radios could be switched between channels, which would mean that only those staff who had their radio switched to that channel could hear transmissions. This was done in circumstances where information was security related and, but not commonly, when a medical emergency occurred.

3.2.3 Prison staff stated that they were aware that information spoken over the radios could be overheard by prisoners and this meant that they would carefully consider what information they communicated. For confidentiality, in an emergency situation they did not use a prisoner’s name, age or the precise nature of the incident. Instead they would only state ‘code red/code blue’ and the location of the incident.

3.2.4 For security reasons the radios were not capable of transmitting outside of the prison. Communicating with the emergency services therefore had to be done via the prison control room, which had an external telephone line.

The control room

3.2.5 The control room is the central hub of the prison and controls or monitors:

• all internal communications
• access through security gates
• CCTV security cameras
• external communication with emergency services
• prison alarms and security systems.

3.2.6 The investigation visited control rooms across the country and observed variability in their staffing levels and environment. All of the control rooms were busy due to the number of simultaneous activities being co-ordinated or observed from them. Control room staffing in different prisons ranged between 2 and 12 personnel, with little difference in the activity being managed within the room. In the control rooms with lower staff numbers the workload described during a medical emergency was difficult to manage. For example, one member of staff would be communicating with staff at the scene of the emergency via the radio and another would make the call to emergency services while standard operations continued.

3.2.7 The control room was a challenging environment, with many distractions such as prison alarms, radio calls, CCTV observations and telephone calls, all adding to a busy task where concentration could be difficult. There were also some control rooms where continuous security alarms were sounding, some of which were ‘false alarms’ that could not be silenced because of faults in the system. The staff in the control room only received limited information about the medical emergency and with all the other distractions it would be difficult for them to retain information being relayed from the scene. There was no provision for extra staff to supplement the control room during a medical emergency and no procedures for reducing the workload.

Situational and clinical information flow and handover

3.2.8 While calls to the emergency services had to be made by the control room some prisons were able to then transfer the calls to a phone in the vicinity of the emergency. This was done so that information from closer to the scene could be relayed to the emergency services and it reduced the burden on the control room. However, the investigation was told that transferring such calls did not always work. This was because the phones were located in offices on the wings, which meant someone needed to be there to answer the call; this was not always the case as the staff would have gone to respond to the emergency. Also, due to the location of the phones, staff relied on the use of a ‘runner’ to relay information between the scene and the person on the phone. Transferring a 999 call to the wing office phone was therefore not common practice across the prisons visited, because although intended to reduce the chain of communication it had been noted to add an extra handover of information between the scene and the emergency services.

3.2.9 For each handover that is introduced into a system the chances of error increase, as shown in a study of emergency care pathways published by the National Institute for Health and Care Research (Sujan et al, 2014). This found that ‘handover failures were particularly likely for patients with longer stays in the ED [emergency department], who received multiple handover’.

3.2.10 The investigation observed different methods of communication between the control room and the prison wings, which included having cordless phones in the wing office. However, prison staff told the investigation that these phones were not being used to patch the emergency services through to the scene of an incident, due to the limitations of the phones themselves. The walls of the wings blocked the signal to the office. In addition, the officers would have to carry the phone with them every time they left the office unattended, which meant there was no guarantee that the phone would be with the first officer at the scene of an emergency. Cordless phones ended up being left in the wing office and the cordless functionality became irrelevant.

3.2.11 Once the healthcare team were present at the scene of an emergency, they assessed the patient and relayed to the control room whether an ambulance was still needed or could be stood down. The control room would then contact the emergency services to update them.

3.2.12 The investigation heard that the information received by the emergency services did not always reflect the medical emergency. This was due to the issue of multiple handovers of information (as described in 3.2.9) and prison staff not understanding clinical terminology. The investigation heard examples such as a patient being described as ‘unresponsive’, which in clinical terms means the patient is unconscious, prompting the dispatch of an air ambulance and multiple ambulance crews. However, on arrival ‘unresponsive’ was found to mean the patient was refusing to answer questions. The ambulance crews and prison staff had different understandings of the word unresponsive, which had significant implications for the emergency response.

3.2.13 On another occasion an ambulance crew had been advised that they were attending a patient aged 90 who had twisted his ankle. Upon arrival the crew found that the patient was a man aged 30 with multiple stab wounds. It was not possible to identify exactly how, or at which point in the communication process, the information changed so much from the scene to the ambulance crew. The crew stated they would have mobilised an air ambulance and a critical care paramedic if they had known the severity of the patient’s injuries.

3.2.14 Ambulance crews described how the lack of accurate and appropriate information directly affected their ability to prepare for callouts to prisons. This included preparing themselves mentally for what they may find at the scene, and planning possible approaches for clinical treatment and equipment they may need to be carried from the ambulance to the scene. This planning is important because of the difficulty of returning to the ambulance for further equipment in a secure environment. Ambulance crews are escorted at all times; the wing doors/gates are locked every time and therefore need to be unlocked and re-locked as the crew are taken through each one. It was not always possible to park the ambulance directly outside the wing, adding to the distance and the number of doors/gates to go through.

3.2.15 Without correct and up-to-date situational information it was not possible for ambulance crews to properly prepare for the incident they were attending. The emergency services rely on accurate information to be able to respond and send appropriate resources, such as air ambulance or rapid response vehicles.

Management of 999 calls

3.2.16 The investigation heard that the number of 999 calls made from each prison varied between 1 and 40 per week. One ambulance trust reported to the investigation that in 2023 they received approximately 5,000 calls from prisons across their region.

3.2.17 All ambulance trusts engaged with during the investigation stated that many of the calls made were stood down before an ambulance was dispatched. These occurred when prison healthcare staff had seen the patient and assessed that an ambulance was not required. Sometimes this would happen during the course of the initial 999 call. Figures supplied to the investigation by ambulance trusts showed that approximately 25% of calls were stood down prior to dispatch.

Ambulance categorisation and dispatch

3.2.18 Ambulance trusts explained that the NHS Pathways triage and assessment system (see 1.6.3) was “not really suitable” to be used for calls from prisons due to the limited information being shared from the prison control room, as discussed in 3.2. Some ambulance trusts used a workaround known as the ‘attend incident function’ which stated:

3.2.19 For the ambulance services that the investigation spoke to, the default categorisation for 999 calls from prisons was category 2 (a target response time of 18 minutes) unless additional information was supplied. The call handlers had the ability to upgrade or downgrade the categorisation; however, the investigation was told that the information needed to do this was rarely available at the time of the initial 999 call.

3.2.20 Prison staff described making follow-up calls to the emergency call centre to update the information about patients. Some prisons had a direct local number for such calls but most made another 999 call to provide this update. This added to the number of calls being made to the service and increased the wait time for the call to be either cancelled, or recategorised.

3.2.21 The investigation was told that crews had responded to calls at prisons that had been incorrectly categorised due to a lack of information. This worked both ways, with some calls being categorised as a category 2 when the patient was not breathing and receiving cardiopulmonary resuscitation (CPR), which would have been classed as a category 1 response if more information had been provided. Crews also described occasions where calls were categorised at category 2 but were not serious enough to meet the criteria for that level of response.

3.2.22 In addition, the investigation was told that ambulance crews would occasionally arrive at the scene and find that they were dealing with multiple patients. They described scenarios in which several patients had overdosed after taking recreational drugs, but the number of patients had not been communicated to the service and that had meant not enough resources had been allocated to the call.

3.2.23 The investigation was told about delays to patient care in prisons and the community due to incorrect categorisation of calls from prisons. This was caused by insufficient or incorrect information being supplied during the initial 999 calls, which in turn affected the categorisation and allocation of ambulances.

Ambulance diversions

3.2.24 The time taken to cancel an ambulance attendance depended on how long it took for the emergency response nurse to arrive at the scene, assess the patient and cancel the ambulance. If it was not cancelled during the initial call, it would be categorised and passed onto the dispatchers.

3.2.25 The experience of ambulance dispatchers needing to divert ambulances to prison calls varied, both between regions and within the same emergency operations centres. Dispatchers described either “regularly diverting ambulances” or having to do it “very occasionally” However, ambulance crews described being regularly diverted to prisons and then being told to stand down and either go back to the call they had originally been responding to or go to a new call.

3.2.26 Crews felt frustrated about this because it meant delays in reaching patients; they felt that diversions regularly led to time being wasted. The investigation obtained data from ambulance trusts on prison calls and whether ambulances were dispatched and attended the prison, or were cancelled while on the way to the prison.

3.2.27 One ambulance service provided the following call figures from prisons for a period of just under 5 months across its region:

• total number of incidents that resulted in cancellation – 1,785
• total number of incidents cancelled before assigning a resource (vehicle and crew) – 1,128
• total number of incidents cancelled before a resource was en route – 26
• total number of incidents cancelled after the resource was en route – 631.

3.2.28 The analysis of these figures shows that of the 1,785 cancelled calls:

• 1,154 incidents were cancelled either during the initial 999 call, while waiting for an ambulance to be dispatched, or after dispatch but before the crew had started their journey to the prison. These were calls that were most likely to have been cancelled by the emergency response nurse after they had arrived on the scene and assessed the patient.
• 631 calls resulted in an ambulance being sent to the prison and the call being cancelled before they arrived. This is what was referred to as a ‘divert’ by the ambulance crews and dispatchers that the investigation spoke to.

3.2.29 The figures were then adjusted to equate to approximate representative figures for a 12-month period. The figures below show the result:

• total number of incidents – 4,284
• total number of incidents cancelled before assigning a resource – 2,707
• total number of incidents cancelled before a resource was en route – 62
• total number of incidents cancelled after the resource was en route – 1,514.

3.2.30 The following analysis has been conducted based on times which ambulance crews told the investigation they experienced when diverting to prisons. Crews described diverts as short as 2 minutes but as long as 10 minutes. Therefore, the investigation has used a 5-minute and a minimum 2-minute divert time for representation. It is important to note that the amount of time spent on calls that are then cancelled are not recorded by any national organisations.

3.2.31 If each of the ‘diverted’ ambulance crews were diverted away for an average of 5 minutes (2 minutes 30 seconds in each direction), which was a common perspective from ambulance crews, then this would result in over 126 hours of divert time. These figures reflected one ambulance trust, and the investigation was told of similar issues by 6 of the 10 ambulance trusts across the country; however, the figures varied across the trusts.

3.2.32 If the figures are representative of the issues experienced by ambulance trusts across the country, then time spent by ambulance crews on unnecessary diversions would equate to approximately 1250 hours of ambulance crew time per year, or over 100 ambulance crew shifts (based on a 12-hour shift pattern).

3.2.33 If the figures were reduced to the lowest amount for the diversions and based on crews only being diverted away for 2 minutes (1 minute in each direction) then this would result in 50 hours and 28 minutes of divert time for the trust and a national projection of approximately 500 hours of crew time. However, ambulance crews described much longer divert times than the short time (2 minutes) used in this analysis. Therefore, the hours spent by crews diverting to prisons is likely to be significantly higher across a 12-month period, in line with the 5-minute estimate.

3.2.34 Ambulance trusts across the country are under pressure and during winter queues of ambulances outside hospitals have been a regular occurrence. This was looked at during the HSIB investigation ‘Harm caused by delays in transferring patients to the right place of care’ (Healthcare Safety Investigation Branch, 2023). Crews told the investigation that the diverts were causing delays to treatment for people who needed it. One ambulance crew told the investigation that “we waste vital time driving to a prison for a call which then gets cancelled and we’ve just added that time on to the response to someone who needs us”.

Improvement of medical emergency response procedures in prisons

3.2.35 There were key areas within the immediate emergency response processes where procedures hindered both the ambulance services and prison service. A significant area for improvement is communication between the organisations about the nature of the emergency, the condition of the patient and what response was being initiated.

3.2.36 Learning has been taken from previous deaths in custody and a rapid prison response system has been implemented. However, without collaborative working with the ambulance services (see 3.4) the procedures for responding to medical emergencies have resulted in unintended consequences as described in this section. The investigation heard examples where the emergency response system hindered the organisations and may have caused delays in care resulting in harm to patients, both inside and outside prisons. Prison staff at the scene of a medical emergency need to be able to communicate directly with ambulance service call handlers, enabling a quick and correct triage and response by the emergency services.

HSSIB makes the following safety recommendation

Emergency vehicle access

3.3.1 Prison vehicle entrances have a ‘vehicle lock’ with two gates (inner and outer) that control vehicles going into and out of the prison. For entry, the outer gate opens and allows the vehicle into a search area. The vehicle cannot go past this point as the inner gate is shut. The outer gate then closes, the vehicle and occupants are searched, and once cleared the inner gate opens, allowing the vehicle to go into the prison. There is a security safety system which does not allow both gates to be open at the same time.

3.3.2 Across the country ambulance staff and prison officers expressed frustration at the medical emergency procedures in relation to ambulance entry and exit. Ambulance crews stated that they regularly encountered delays at the front gate of prisons while trying to get through the vehicle lock. Prison officers stated that ambulance crews were often not prepared for the required security measures, such as vehicle search and handing in mobile phones.

3.3.3 The investigation observed vehicles accessing prisons, noting the time taken for the vehicle gates to open and close (approximately 20 to 25 seconds depending on the prison), and how long the search process took for a trade vehicle (approximately 7 to 8 minutes). Prison staff told the investigation that ambulance searches were a lot quicker and they could be through the vehicle lock in “a couple of minutes”. From the observations undertaken by the investigation these times were unachievable if the security procedures were followed. The opening and closing of gates alone took close to the quoted 2-minute notional time.

3.3.4 During an emergency there is a lot of activity in a prison’s gate area, including:

• opening and closing both the vehicle gates
• escorts preparing to take the ambulance to the wing
• confirming the identity of the ambulance crew
• securing the ambulance crew members’ mobile phones and vapes
• searching the ambulance, if quickly
• logging the ambulance into the gate entry log
• communicating with the prison via radio about progress.

3.3.5 The level of activity would likely result in staff being unaware of how long the process takes. The investigation heard from ambulance crews that it was often apparent that the process of getting the ambulance through the vehicle lock was slow, with crews being asked questions about the equipment they may have on board. The investigation heard of ambulance crews being asked if they had any sharps (needles) or drugs (medication) on board, which all ambulances and rapid response vehicles carry. This added to the frustration of crews trying to access the prison and the time taken to reach the patients.

3.3.6 Some prison staff described a process for getting ambulances through the vehicle lock that did not match with the prison policy. The policies being applied by prisons varied across the country, with local prison policies not always matching ambulance trust policies, such as whether ambulance crews should hand over their mobiles phones on entry to the prison. The local prison policies, known as local security strategies, varied to the point that it was not possible to determine before they arrived what the security requirements may be for the ambulance crews.

3.3.7 In addition, prison staff told the investigation that ambulance crews were inconsistent in their willingness to comply with their security policies; this was also reflected in statements made by ambulance crews. Some crews were ready and prepared to hand over their personal phones, whereas other crew members told the investigation that they were not prepared to hand these over because they had ‘work apps’ on them. The inconsistency added to the frustration that both prison staff and ambulance crews felt.

Ambulance service attendance information

3.3.8 Prison staff from the control rooms informed the investigation that when they made a 999 call they would be given limited information by the ambulance service about the response. Staff said that they “mostly got an ETA [estimated time of arrival]”, “but not always” and that they rarely got information about what vehicles were being sent to the prison.

3.3.9 The most common response to prison emergency calls was a double-crewed ambulance (an ambulance with 2 staff on board). However, depending on the scenario this may be increased with more ambulances, rapid response vehicles or an air ambulance. Ambulance service policies stated that the following information would be passed from the ambulance service to the prison:

• the ETA
• vehicle registration number – of all vehicles attending
• vehicle call sign – of all vehicles attending
• names (some include personal identification numbers) of staff attending.

3.3.10 Prison staff told the investigation that they were often given an ETA for an ambulance, which was the most common vehicle to respond, but they were rarely given any other details, such as how many vehicles were coming. Ambulance service staff said that they gave an ETA if they knew it, but that this was not always possible if the crew was not dispatched immediately. This had a direct impact on the readiness of the prison staff at the gate as they did not always know when an ambulance was arriving and could postpone other procedures if the ambulance ETA was 2 hours. The other information, such as vehicle registration numbers, was not routinely passed on by dispatchers to callers and would therefore be a bespoke procedure for prisons.

3.3.11 The investigation was also told of many scenarios where ambulances had arrived at prisons and been unable to gain immediate access to the vehicle lock because another vehicle was being processed through. The gate staff stated that this was because an ambulance or rapid response vehicle had already been cleared through and they did not know a second vehicle had been sent, and therefore had not kept the vehicle lock clear.

3.3.12 The lack of information being passed to the prison from the emergency call centre, particularly in respect of multiple vehicles responding to a call, meant that the staff at the gate were unable to prepare in advance for the vehicles that were being sent, by:

• preparing the entry into the prison access log
• getting vehicle escorts ready to escort the emergency vehicles to the scene
• keeping the vehicle lock clear to ensure immediate access for all emergency vehicles attending.

Emergency vehicles’ exit from the prison

3.3.13 The investigation was told of delays to ambulances being released from prisons, mainly caused by prison paperwork not being ready. The paperwork included a prisoner risk assessment (completed by the prison and healthcare staff) and the prisoner escort record form (completed by prison and healthcare staff), which includes healthcare details about the patient being taken to hospital. The investigation heard different accounts of the length of time it took to complete these forms. The need to arrange prison staff escorts occasionally led to delays, with prisons having difficulty finding staff who could be released to conduct the escort.

3.3.14 Variation was observed in processes across the country for releasing ambulances with patients on board. Some prisons would conduct a longer search of the ambulance on the way out than they would on entry to the prison. However, some prison staff stated that they would not delay the ambulance at all because “the patient getting to hospital quickly is all that matters”. The difference in quoted times for clearance of the vehicle lock on exit ranged from 2 minutes to 10 minutes. The variation in timings and processes between prisons was observed in nearly every prison the investigation visited.

3.3.15 Guidance from HMPPS (PSI 03/2013) did not set out specific processes for emergency vehicles accessing or exiting prisons, instead giving a general overview of the intended outcome and allowing individual prisons to write their own procedures. There was also no specific process guidance at a national level for ambulance services to follow, again resulting in local-level variation.

3.3.16 The investigation also heard differing understanding of policies from ambulance crews, particularly about handing over mobile phones and other items when entering a prison. Confusion arose about whether personal phones with work apps, such as the Joint Royal Colleges Ambulance Liaison Committee app, were classed as work phones and therefore whether crew members could keep them upon their person while in the prison. Ambulance crews described delays while discussions took place with the prison gate staff about the handover of devices. Prison gate staff and ambulance crews told the investigation about the frustration the subject had sometimes caused.

3.4.1 Prison officers told the investigation that they had no way of raising concerns or issues that arise during emergencies. Prison management said that they were aware of incidents and some problems that had occurred, but they had no communication with ambulance trusts, other than the attending crew, to address these.

3.4.2 When speaking with the investigation, ambulance crews expressed frustration that the same problems were being encountered repeatedly. The variation between prison security procedures and the information they were given about the patients, for example, could not be addressed as there was nothing they could do other than raise an incident report. They stated that this had never resulted in any improvements across the ambulance and prison services response to medical emergencies.

3.4.3 There was evidence from both ambulance crews and prison staff that areas for improvement were being missed. All staff that the investigation spoke with were keen to engage and expressed a desire to hear of problems from the other perspective and to be able to enact change.

3.4.4 An area for improvement that was expressed by prison staff and ambulance crews was in the formation of policy. The organisations’ policies were written in isolation of each other and therefore did not account for each other’s needs – for example, the issue of mobile phones being handed in at the prison gate. These policy conflicts had directly resulted in delays in ambulance crews accessing prisons or getting to the patient once inside the prison.

3.4.5 The investigation heard from ambulance trusts and prisons that the ability to feed back to each other would be important for them to be able to reflect on issues that arise. There were some examples where ambulance trusts had started to try to work together, but these were based on informal relationships and were not spread across the whole region. With no formal mechanism in place for this, neither side was able to raise concerns or initiate change to make the medical emergency response more efficient, collaborative and safer for patients.

HSSIB makes the following safety recommendation