Investigation report: Oxygen issues during the COVID-19 pandemic

Concerns about patient safety in any area of NHS-funded healthcare in England can be referred to us by any person, group or organisation. We review these concerns against our investigation criteria to decide whether to conduct a national investigation. National investigation reports are published on our website and include safety recommendations for specific organisations. These organisations are requested to respond to our safety recommendations within 90 days, and we publish their responses on our website.

We investigate incidents in NHS maternity services that meet criteria set out within one of the following national maternity healthcare programmes:

• Royal College of Obstetricians and Gynaecologists’ ‘Each Baby Counts’ report
• MBRRACE-UK ‘Saving Lives, Improving Mothers’ Care’ report.

Incidents are referred to us by the NHS trust where the incident took place, and, where an incident meets the criteria, our investigation replaces the trust’s own local investigation. Our investigation report is shared with the family and trust, and the trust is responsible for carrying out any safety recommendations made in the report.

In addition, we identify and examine recurring themes that arise from trust-level investigations in order to make safety recommendations to local and national organisations for system-level improvements in maternity services.

For full information on our national and maternity investigations please visit our website.

This investigation describes an emerging safety risk where there has been an increased demand for oxygen on hospital wards during the COVID-19 pandemic.

As an example, which is referred to as ‘the reference event’, the investigation explores an incident where an acute hospital trust (the Trust) declared a major incident when demands on its oxygen supply, delivered via its medical gas pipeline system (MGPS), led to patients being diverted to different hospitals, elective (planned, non-emergency) surgery being cancelled, and a need to reconfigure ward environments.

COVID-19 can cause severe inflammation of the lungs, which affects people’s ability to breathe and get enough oxygen into their bloodstream. As a result, an increased number of patients with COVID-19 have required oxygen therapy. Insufficient oxygen supply to seriously ill patients can have very severe consequences, including death.

The investigation explored the role and understanding of MGPS in the response to the COVID-19 pandemic, focusing on current engineering and technical systems that help to ensure MGPS operate effectively. The investigation acknowledges that lack of oxygen supplies has implications for the clinical management of patients but has not explored the clinical management of patients requiring oxygen therapy in this report.

This investigation’s findings, safety recommendations and safety observations aim to improve the management and understanding of MGPS to improve care for patients across the NHS.

During the first wave of the COVID-19 pandemic in spring 2020, the Trust had identified where patients requiring oxygen therapy could be cared for within the hospital to ensure demand on the MGPS was evenly distributed. It prepared a plan that accounted for the anticipated demand on oxygen supplies and the type of oxygen therapy it expected to provide to patients at that time.

In autumn 2020, the Trust found that two patients being provided with oxygen therapy in the high dependency unit (HDU) had desaturated (lost the required levels of oxygen from their blood). It was unclear why the patients had desaturated, and the Trust considered whether there had been a failure in the MGPS within the HDU. As a precaution, the Trust moved patients from the HDU to the operating theatres where sufficient oxygen supply was available.

The Trust initially diverted incoming patients to another of its hospital sites and cancelled all elective surgery while further investigation took place. However, oxygen demand at its other sites also began to cause concern and so the Trust declared a major incident and requested all incoming patients be diverted to other trusts. The Trust began planning further patient moves and identified mitigating measures that would allow it to carry out a test on the MGPS in the HDU.

The Trust carried out a pressure test on the MGPS to identify any failures. The test results identified several areas where there was a reduced ability to provide the anticipated flow of oxygen from the MGPS.

The Trust initially understood that the HDU should have been capable of providing a specified flow of oxygen and had used this to calculate the number of patients that could be treated, and type of oxygen therapy that could be used. However, the anticipated flow of oxygen to the HDU, and other wards, was likely impacted by excess earlier demand for oxygen from the MGPS system prior to the pipeline reaching the HDU.

The Trust reported that the overall demand for oxygen on the MGPS never exceeded more than 56% of its total capacity, meaning that the Trust had spare capacity to generate oxygen. The Trust investigation identified that any possible lack of oxygen flow available to the HDU was instead due to limitations in its existing pipework infrastructure, the distribution of patients across the hospital, and the types of oxygen therapy required to treat patients across the hospital.

The Trust considered that the plan it had developed in spring 2020 had not been effective due to a range of factors, including:

• a lack of COVID-19 patient demand on the MGPS during the first wave of the pandemic meant that the problems caused by additional demand were not identified
• the need to group patients together based on their clinical need, creating uneven demand on the MGPS and delivery of oxygen to the bedside
• the baseline figures used to calculate potential demand on oxygen consumption to treat COVID-19 patients being too low
• variation in the volume of oxygen used by different equipment capable of delivering oxygen therapy
• the existing MGPS pipework infrastructure
• general demand for bed spaces caused by the number of patients requiring treatment for COVID-19.

The Trust carried out work to revise its plans and understand how this impacted on where patients requiring oxygen therapy could be cared for within the hospital to maintain sufficient oxygen supply across the system. This work enabled the Trust to redistribute patients across the hospital to provide a more equal spread of oxygen demand and allowed one of its hospital sites to readmit patients.

Following this work, and a decrease in the number of patients being treated at both hospital sites due to the impact of diverting patients to other hospitals, the Trust was able to stand down from the major incident and start to admit patients to both hospital sites, seven days after the major incident was declared.

The two HDU patients that had become desaturated recovered their oxygen saturation levels without any adverse effects. The investigation has been unable to conclusively determine whether the desaturation was associated with any pressure loss or reduction in oxygen flow within the HDU or other factors involved in their care.

HSIB identified the incident via the NHS Strategic Executive Information System (a national database of serious incidents in healthcare). HSIB engaged with the Trust to explore the circumstances of the incident report and this formed the basis of the reference event.

HSIB engaged with national organisations to understand the safety risk across the NHS and to learn more about the current design, management and regulation of MGPS. National stakeholders told the investigation they would be eager to ensure that learning from the COVID-19 pandemic was captured to help enhance the operation of MGPS going forward.

• The MGPS is a critical system in the safe and effective operation of an acute NHS hospital.
• The COVID-19 pandemic placed unprecedented demands on MGPS due to the number of patients requiring oxygen therapy and the different types of oxygen therapy equipment used.
• There is a lack of shared ownership and knowledge of MGPS among hospital-based multidisciplinary teams; this limits trusts’ ability to effectively respond to MGPS patient safety concerns.
• Organisations that utilised a multidisciplinary approach to understanding and planning the MGPS involvement in the COVID-19 response, including the impact of ward moves and the choice of oxygen therapy, were better able to respond to demands on the MGPS system.
• Current guidance on the design and management of MGPS contained within the relevant health technical memorandum is outdated and does not reflect developments in oxygen therapy and challenges in managing MGPS.
• The investigation acknowledges work carried out by national bodies to help NHS trusts address the emerging risks to MGPS, including the issue of interim guidance in addition to the HTM and capital investment in MGPS.
• Current assurance mechanisms for MGPS are not effective in ensuring that MGPS-related patient safety concerns are proactively identified and resolved.
• A lack of financial investment in updating MGPS infrastructure created challenges for NHS trusts in responding to the COVID-19 pandemic.

In addition, HSIB has released two interim bulletins during the investigation highlighting key findings emerging from the initial reference event investigation and the role of medical gas committees: ‘Oxygen issues during the COVID-19 pandemic. January 2021’ and ‘Oxygen issues during the COVID-19 pandemic. March 2021’.

1.1.1 Oxygen is widely used in healthcare settings and is classified as a medicine in the UK (British National Formulary, n.d.). The British Thoracic Society (BTS) guideline (2017) suggests oxygen is probably the most common medicine used in the care of patients who visit hospital due to medical emergencies. It identifies three main reasons for administration of oxygen in hospitals:

• to correct hypoxaemia (low levels of oxygen in a patient’s blood)
• to prevent possible hypoxaemia
• to alleviate breathlessness.

1.1.2 In the treatment of COVID-19, oxygen is a key medicine that aims to help patients maintain the oxygen levels in their body.

1.1.3 In hospital, there are two main ways of providing oxygen to patients who need it: a medical gas pipeline system (MGPS), which is integrated into the hospital buildings, or portable oxygen supplies that use cylinders.

1.1.4 HSIB has previously reported on specific safety risks associated with both the use of piped oxygen (Healthcare Safety Investigation Branch, 2019) and portable oxygen supplies (Healthcare Safety Investigation Branch, 2018). This investigation focused on piped oxygen supplies.

Oxygen therapy equipment

1.1.5 Oxygen therapy can be provided by a range of medical devices, including:

• ventilators, which are machines that pump oxygen-enriched air into patients’ lungs – a process known as ‘ventilation’
• continuous positive airway pressure (CPAP), where a machine delivers constant and steady positive air pressure to a patient via a mask or nose piece
• high flow nasal oxygen, which is delivered to a patient via a nose piece at higher rates of flow than that delivered by traditional oxygen delivery devices.

Each device typically mixes pure oxygen with atmospheric air to provide oxygen therapy to a patient.

1.1.6 To be effective, each different form of treatment may require pure oxygen to be provided at different flow rates. Typically, this can range from ventilators that are capable of operating at 15 litres per minute (L/min), CPAP which may be provided from around 30L/min, and high flow nasal oxygen which may require more than 60L/min, all dependent on the individual patient’s requirements.

1.2.1 The main components of an oxygen MGPS are:

• Vacuum insulated evaporator (VIE): a pressurised vessel that stores liquid oxygen (see figure 1).
• Vaporisers: which convert liquid oxygen into oxygen gas (see figure 2).
• Regulator control panel: which regulates the flow of oxygen gas from the VIE into the hospital building.
• Master and local alarm panels: which warn of potential issues with the supply of oxygen gas.
• Associated pipework that transports the oxygen gas around the hospital.

1.2.2 These components work together to deliver oxygen gas safely into the hospital environment. Each MGPS provides a specified flow of oxygen gas to a hospital (commonly measured in available L/min) at a specified pressure; this should maintain the desired flow.

1.2.3 The investigation acknowledges that the technical language used to refer to different aspects of oxygen delivery may cause confusion for non-specialists. Such technical terms include:

• pressure (the force which is applied to pushing the oxygen through the pipes, regardless of the size of the pipe)
• volume (the quantity of oxygen that can be moved through a pipe; this is completely dependent on the size of pipe)
• flow (how much oxygen can be delivered over a given time; this is expressed in the report as L/min)
• capacity (the maximum amount of oxygen that the system is able to produce; this is expressed in the report as L/min).

To avoid overly technical language, this report refers broadly to ‘flow’ to indicate the movement of oxygen in the MGPS and the delivery of oxygen to the patient bedside. Where technical terms are required to understand more about the context being discussed, these are used with reference to the definitions set out above.

1.2.4 Typically, MGPS are arranged as either:

• A radial system (see figure 3): a pipeline system which begins at the piped oxygen source and ends at another point so that every outlet has only one possible route of oxygen supply.

• A ring main (see figure 4): a pipeline system which begins and ends at the piped oxygen source so that every outlet has two possible routes of oxygen supply.

1.2.5 HSIB understands that many NHS hospitals operate on radial systems. In radial systems, oxygen being drawn off the system earlier in its progress through the hospital reduces the potential pressure available to oxygen outlets later in the system. The location of patients who need oxygen therapy, and type of oxygen therapy being provided, within the hospital needs to be carefully planned to avoid unequal pressure on the MGPS and unanticipated reductions in oxygen supply at other points in the system.

1.2.6 HSIB consistently heard of an analogy to a garden hosepipe when specialist MGPS users were trying to explain MGPS demand to non-specialist users. This may be helpful for a non-specialist reader. In the hosepipe analogy, there is a sprinkler attached to a hosepipe spraying water. The more sprinklers you attach to the same hosepipe the lower the pressure is to each sprinkler until only some, or none, of the sprinklers are working as expected.

Health technical memorandum 02-01: medical gas pipeline systems

1.2.7 Health technical memoranda (HTM) are intended to give comprehensive advice and guidance on the design, installation and operation of specialised building and engineering technology used in the delivery of healthcare. The responsibility for the production and management of HTMs transferred from Department of Health and Social Care to NHS England and NHS Improvement (NHSE/I) in February 2017.

1.2.8 HTM guidance is applicable to new and existing hospital facilities and provides best practice engineering standards and policy. New hospital facilities, and facilities that are being refitted, should be constructed in line with any current HTM guidance at the time of construction.

1.2.9 Guidance on the installation and management of MGPS is set out in ‘Health technical memorandum 02-01: medical gas pipeline systems’, parts A and B (Department of Health, 2006). NHSE/I told the investigation that in 2017 it reviewed all the health technical memoranda and health building notes and identified that 18 required updating. The HTM was one of these documents and work began to update the HTM in late 2019. However, due to the demands placed on the healthcare system during the COVID-19 pandemic, work to update the HTM had needed to be placed on hold. As an interim measure, further guidance was provided to NHS trusts to help manage the MGPS response to COVID-19.

1.2.10 Part A covers the design and installation of MGPS and part B covers their operational management. Part B defines a number of individual roles and governance processes that operate to ensure the safe and efficient use of the MGPS. These include:

• The medical gas committee (MGC): a forum to oversee the general operation and management of the MGPS.
• Authorising engineer: a suitably qualified, independent MGPS specialist who provides a report on the MGPS system, recommends the appointment of authorised persons, and may deliver specialist training.
• Authorised persons: who are responsible for the day-to-day management of the MGPS at a particular site or sites.
• Designated nursing and medical officers who liaise from a clinical perspective on any matters affecting the MGPS.

1.2.11 The MGPS is designed to achieve set flow rates in accordance with the HTM. For a standard hospital ward, the HTM provides for a minimum of 1 in 4 patients to receive 6L/min of oxygen and one bed requiring 10L/min of oxygen. This means that a 28 bedded ward designed in accordance with the HTM would have a minimum expected flow rate of 50.5L/min. For a high dependency unit (HDU), the HTM provides for a minimum of one bed to receive 10L/min of oxygen and three quarters of the remaining beds to receive 6L/min. This means that a seven bedded HDU designed in accordance with the HTM would have a minimum expected flow rate of 37L/min.

1.2.12 Once these flow rates are exceeded the redundance (extra flow) in the system is used to support additional usage, until the maximum flow of the pipelines is reached. At this point, there may be a drop in pressure and loss of flow.

1.2.13 The majority of hospital wards and departments have had to accommodate higher than designed flow rates to respond to the COVID-19 pandemic. This has sometimes placed demand on the MGPS to deliver more than it may have been designed for.

1.2.14 In total, the HTM is 336 pages long and contains a comprehensive overview of all aspects of MGPS. The investigation has not completed a full review of all elements of the HTM and has focused on specific elements of the HTM relevant to this investigation. NHSE/I told the investigation that it would look to incorporate evidence based learning from the pandemic into any revised HTM document.

NHS Premises Assurance Model

1.2.15 The NHS Premises Assurance Model (PAM) was first published in 2013 and has been regularly updated since that time (NHS England and NHS Improvement, n.d.). It is intended to support acute NHS trusts in assuring their estate is safe, efficient, effective and of a high quality.

1.2.16 The NHS PAM is intended to support a multidisciplinary approach to understanding the views of the hospital staff that use the estate and its facilities. The PAM includes assurance questions on MGPS.

1.3.1 The Central Alerting System (CAS) is a web-based system for issuing patient safety alerts, important public health messages and other safety-critical information and guidance to the NHS and other organisations, including independent providers of health and social care (Medicines and Healthcare products Regulatory Agency, n.d.).

1.3.2 The CAS is funded and operated by the Medicines and Healthcare products Regulatory Agency (MHRA). In operating the CAS, the MHRA works closely with all organisations that make up the National Patient Safety Alerting Committee (NaPSAC).

1.3.3 Alerts available on the CAS website include:

• national patient safety alerts (from NHSE/I and MHRA)
• NHSE/I estates and facilities alerts
• Chief Medical Officer alerts
• Department of Health and Social Care supply disruption alerts.

Oxygen-specific safety alerts

1.3.4 On 31 March 2020, NHSE/I issued an estates and facilities alert about the use of high flow oxygen devices (NHS England and NHS Improvement, 2020a) via the CAS. NHSE/I identified that a far greater proportion of patients would require oxygen therapy and ventilation because of the COVID-19 pandemic.

1.3.5 The alert warned that ‘if the demand [for oxygen] through multiple wall outlets exceeds the maximum capacity of the VIE delivery system, there is a risk of a rapid pressure drop in oxygen supply pipes. This could lead to a failure of oxygen delivery systems throughout the hospital. There is also a risk of rapid and unpredictable depletion of the VIE. Both of these situations present a potentially significant risk to multiple patients simultaneously’ (NHS England and NHS Improvement, 2020a). The alert asked NHS trusts to take a range of actions to mitigate against this risk.

1.3.6 On 6 April 2020, NHSE/I issued a further estates and facilities alert on oxygen usage (NHS England and NHS Improvement, 2020b) via the CAS. NHSE/I identified that, with hospitals treating a large proportion of their inpatients with oxygen for COVID-19, some hospitals were drawing more oxygen than the maximum flow for which their MGPS were designed. The report identified that this ‘carries the risk of icing [the build-up of ice on MGPS pipework] that could cause flow to drop unexpectedly, compromising supply to patients and/or permanent damage to the system’. The alert asked NHS trusts to ‘engage their engineers, with clinicians and managers, to safely look after their patients and plan their surge capacity’.

1.3.7 On 19 November 2020, NHSE/I issued a further estates and facilities alert on oxygen supply and fire safety (NHS England and NHS Improvement, 2020c) via the CAS. NHSE/I reminded NHS trusts to be aware of the risks set out in the estates and facilities alert issued on 31 March 2020.

1.3.8 The alert identified that demand for oxygen during the winter was likely to be higher than in spring due to increased general activity in NHS trusts since that time, and the continuing operation of operating theatres and recovery areas that may have been inactive during the response to the first wave of the pandemic.

1.3.9 The alert also acknowledged the potential impact on oxygen supplies of admitting large numbers of COVID-19 patients, creating new patient bays or wards to care for patients with COVID-19 or other respiratory illnesses, and the potential impact of older hospital estates.

1.3.10 The alert asked trusts to ‘establish leadership teams including key clinical leaders and hospital oxygen engineering teams (including expertise from medical gases committees) to ‘ensure oxygen demand does not outstrip supply’ (NHS England and NHS Improvement, 2020c). NHS trusts were also asked to carry out a range of further mitigating actions.

NHS England and NHS Improvement COVID-19 guidance on piped oxygen systems

1.4.1 In August 2020, NHSE/I issued guidance about the actions necessary to mitigate the effect of the COVID-19 pandemic on the performance of oxygen systems (NHS England and NHS Improvement, 2020d).

1.4.2 The guidance was issued in response to concerns that there were weaknesses in equipment, system configuration and operational management in responding to the increased pressure caused by COVID-19. None of the issues identified included any concern that liquid oxygen supplies may be exhausted.

1.4.3 The guidance incorporated a number of key messages and tasks aimed at assisting hospitals to understand and address the potential limitations of their MGPS and how this may impact on oxygen provision to patients.

NHS Specialist Pharmacy Service guidance

1.4.4 The NHS Specialist Pharmacy Service (SPS) is commissioned by NHSE/I to ‘improve the use of medicines to help people live longer, fuller lives’ (NHS Specialist Pharmacy Service, n.d.).

1.4.5 In January 2021, NHS SPS produced a range of resources (NHS Specialist Pharmacy Service, 2021a) directed toward pharmacy teams to assist in the response to the COVID-19 pandemic and address the use of oxygen, as this is classed as a medicine. These included guidance documents and resources covering:

• chief pharmacist responsibilities for medical gases
• MGPS
• governance processes for medical gases
• use of medical gases and oxygen housekeeping
• medical gas committees.

Oxygen saturation

1.4.6 Oxygen saturation is the measure of oxygen available in the blood stream and is measured as a percentage. A person’s body requires a specific amount of oxygen in the bloodstream in order to function effectively. Typically, a normal range of oxygen saturation for a healthy adult is greater than 94%. Anyone with oxygen saturation below this level may require support with oxygen therapy with the aim of increasing their oxygen saturation level.

1.4.7 On 9 April 2020, NHSE/I issued guidance to NHS trusts on the optimal use of oxygen therapy during the coronavirus pandemic (NHS England and NHS Improvement, 2020e). The guidance recognised that as the number of patients in hospital with COVID-19 increased, there would be increased demand on the flow of oxygen delivery within hospitals.

1.4.8 To support prioritisation of oxygen flow for the most severely ill patients in hospital, the guidance set out that:

• oxygen prescribing targets for all adults treated in NHS hospitals should be adjusted from the current range (of oxygen saturation 94% to 98%) to oxygen saturation of 92% to 96% in the first instance
• a target range of 90% to 94% may be considered if clinically appropriate by hospitals according to prevailing oxygen flow demands.

1.4.9 This investigation did not consider issues relating to the clinical decisions about providing oxygen to patients for treatment of COVID-19. However, the need to issue this guidance highlights the direct impact that inadequate oxygen supply via MGPS can have on patient care and clinical decision making.

3.1.1 HSIB identified a reference event where an acute hospital trust declared a major incident when demands on its MGPS led to patients being diverted to different hospitals, elective surgery being cancelled, and a need to reconfigure ward environments. The Trust had sufficient supplies of liquid oxygen but its MGPS was unable to deliver the flow of oxygen gas required to meet all patient needs.

3.1.2 COVID-19 can cause severe inflammation of people’s lungs, affecting their ability to breathe. As a result, an increased number of patients have required oxygen therapy within hospitals which has placed increased demand on MGPS. Insufficient oxygen supply to seriously ill patients can have very severe consequences, including death.

3.1.3 The reference event highlighted a national safety risk where MGPS may be unable to meet their anticipated output due to factors related to the COVID-19 pandemic.

3.2.1 HSIB conducted an initial scoping investigation which determined that the patient safety concern met the criteria for investigation (see below). HSIB’s Chief Investigator authorised a national investigation.

Outcome impact – what was, or is, the impact of the safety issue on people and services across the healthcare system?

An inability to provide appropriate flow of oxygen to patients poses significant risks to patient health and wellbeing. Where an insufficient volume of oxygen is available it may be necessary for clinicians to manage decisions about how available oxygen supply should be used. This may require patients to be prioritised for care, or for target oxygen saturation levels (the percentage of oxygen in the blood) to be lowered to increase the number of patients that can be cared for with the available oxygen supply (NHS England and NHS Improvement, 2020). This can have significant impacts on patient care.

Systemic risk – how widespread and how common a safety issue is this across the healthcare system?

Oxygen is widely used in healthcare settings and is one of the most common medicines used in the care of patients who need emergency treatment. Incidents reported to the NHS Strategic Executive Information System (StEIS) and national news articles have drawn attention to challenges hospitals have faced in providing sufficient oxygen to patients due to the increased demand on MGPS.

HSIB is aware, via national media, NHS reporting and other information sources, of at least a dozen NHS trusts where concerns were raised. These reports created significant concern to healthcare workers and patients about the ability of MGPS to operate effectively in the NHS response to COVID-19 and caused a range of trusts to take emergency action to mitigate against this risk. The hospitals reporting pressures on MGPS varied in size, age of hospital estate and geographical spread across England.

In each reported instance that HSIB is aware of, sufficient liquid oxygen supplies have been available to trusts. However, there have been limitations in the ability of MGPS to meet the volume of oxygen gas required to treat all patients that may require oxygen therapy in the preferred clinical area of the trust.

Learning potential – what is the potential for an HSIB investigation to lead to positive changes and improvements to patient safety across the healthcare system?

Local action has been taken by a number of trusts that have encountered oxygen supply issues. In addition, estates and facilities safety notifications, in-depth guidance and support have been made available by NHS England and NHS Improvement (NHSE/I) (NHS England and NHS Improvement, 2020a; 2020b; 2020c) and the NHS Specialist Pharmacy Service (2021a) to address and mitigate potential issues with MGPS. Despite these measures, issues with oxygen provided via MGPS continue to be reported.

An HSIB national investigation can consider the systemic safety issues relating to the provision of oxygen via MGPS and examine why existing measures aimed at addressing the risk to MGPS may not have been fully effective. The investigation can inform any future response to COVID-19 or other respiratory viruses, which continue to be recognised as a significant factor that could lead to future global pandemics.

3.3.1 The investigation was completed between January and May 2021 as a priority in response to the emerging resurgence of the COVID-19 pandemic during this period.

3.3.2 The HSIB investigation team members were from a range of backgrounds including:

• healthcare systems
• human factors investigation
• engineering
• clinical and non-clinical subject matter advisors (SMAs).

3.3.3 A range of evidence was gathered and reviewed by the investigation at the reference event Trust, including:

• all relevant Trust MGPS documentation
• interviews with Trust staff, including representatives from estates and facilities, medical engineering, clinical directorates, and operations
• observations of the Trust’s MGPS and working practices.

3.3.4 The investigation engaged with a further four NHS trusts to gather information about local experiences of the use of MGPS during the COVID-19 pandemic. For each trust, this included collating MGPS documentation, interviews with clinical and estates and facilities staff, and observations of MGPS.

3.3.5 Stakeholders across the healthcare system were identified to seek their perspectives on the challenges placed on piped oxygen supply during the COVID-19 pandemic. The investigation analysed national guidance and documentation and conducted interviews and conversations with clinical, pharmacy, regulatory, estates and facilities, and commercial representatives from:

• Department of Health and Social Care
• NHSE/I
• NHS Specialist Pharmacy Service
• Care Quality Commission
• Medicines and Healthcare products Regulatory Agency
• Royal Pharmaceutical Society
• British Compressed Gas Association
• British Thoracic Society
• Institute of Healthcare Engineering and Estates Management.

3.3.6 The evidence gathering process adopted an iterative approach; as further information was gained, additional sources were identified. HSIB considers it has struck an appropriate balance by providing information that may be beneficial to the various layers of the healthcare system in a timescale that would enable a practical response to any further peak in COVID-19 activity.

3.3.7 The findings were shared with the stakeholders identified by the investigation. This enabled checking for factual accuracy and overall sense-checking. The stakeholders contributed to the development of the safety recommendations based on the evidence gathered.

3.4.1 HSIB does not seek to apportion blame or liability in its investigations. It considers the healthcare system in its entirety to identify the factors that have contributed to the reference event. A range of methods were used to collect and analyse evidence during the investigation.

3.4.2 The Systems Engineering Initiative for Patient Safety (SEIPS) model (Carayon et al, 2006) was used as the overarching framework for the investigation. The SEIPS framework provides a way to understand the relationship between the structures, processes, and outcomes of healthcare systems.

3.4.3 In addition, the SHELL model (Skybrary, 2019) was used as a basic aid during visits to the reference event trust and observation sites to help direct evidence collection, and the Systems Theoretic Accident Model and Process (STAMP) model (Leveson, 2012) was considered to aid in the understanding of the control structures and feedback mechanisms in place for MGPS.

Shared understanding of MGPS

The reference event

4.1.1 The Trust told the investigation that at the beginning of the pandemic it was conscious of the potential impact of the risk around its MGPS. Because of this it assigned its clinical, estates and medical devices teams to work together to develop the red dot plan. This helped to encourage a shared understanding of the potential challenges that COVID-19 could cause to the MGPS and for shared decision making on how to respond in advance of the first wave of COVID-19.

4.1.2 Staff told the investigation that the Trust was not heavily impacted by the first wave of the COVID-19 pandemic in spring 2020. This meant that the red dot plan was never fully tested by a substantial increased demand on the MGPS. Due to operational demands, it was also not possible to simulate demands on the MGPS or carry out more specific testing as this would have affected the delivery of patient care.

4.1.3 However, staff told the investigation that “wave one looked very different to wave two”. In wave two (autumn 2020), the Trust had re-established elective activity and Trust staff told the investigation that general hospital activity levels had also increased. This meant that, in addition to an increase in COVID-19 activity, the Trust was also using its MGPS in the care of other patient groups who were not being cared for in hospital during wave one. The changing demands on the type of oxygen therapy used for the treatment of patients with COVID-19 also created significant additional demand on the MGPS.

4.1.4 The Trust was not able to rely on similar processes of shared decision making that helped it to develop the red dot plan to respond to emerging considerations in its response to wave two. The pressures faced by NHS trusts during the pandemic and the impact of COVID-19 on hospital attendance, staff and resources during this period have previously been considered by HSIB (Healthcare Safety Investigation Branch, 2020).

4.1.5 Due to this lack of shared decision making, staff at the Trust explained that there was no shared understanding among different staff groups about how the MGPS worked, as knowledge about how to respond to COVID-19 changed and placed different demands on the MGPS. Factors that impacted on this lack of shared understanding are considered below.

4.1.6 Estates teams understood the technical aspects of how oxygen was delivered to the patient bedside and the anticipated demands that would be placed on the MGPS system by COVID-19 in line with the initial understanding of how COVID-19 may be treated – that is, primarily by patients being placed on ventilators. However, they did not understand the changing clinical context and treatment options being used by clinicians as the pandemic progressed, which created additional demand on the MGPS. This was seen in the change from an expectation that patients would be put onto ventilators to patients receiving other forms of oxygen therapy requiring greater oxygen usage.

4.1.7 Medical engineering teams understood the specific requirements of different devices that could deliver oxygen therapy. This included variations in the amount of oxygen capable of being used by different machines, the mix of oxygen and atmospheric air that machines used to provide care, and the points at which equipment would become ineffective if inadequate oxygen pressure was provided.

4.1.8 Clinical and pharmacy teams understood the patient requirements for oxygen therapy and the target oxygen saturation levels required. This included an acknowledgement of the changing target oxygen saturation levels for patients, in line with NHS England and NHS Improvement (NHSE/I) guidance (NHS England and NHS Improvement, 2020e). However, they did not necessarily understand the different oxygen requirements of different pieces of equipment and the potential impact on the MGPS of clinical decisions about oxygen therapy.

4.1.9 There are competing priorities for clinical and non-clinical teams in responding to the COVID-19 pandemic. Estates teams were conscious that demands on the MGPS meant that oxygen flow needed to be balanced across the hospital estate to ensure adequate supply was available for patient care. If considered as an estates issue in isolation, then an ideal solution would have been to place patients in multiple areas of the hospital to balance the MGPS and maintain sufficient supply.

4.1.10 Clinical teams were conscious of the infectious nature of COVID-19 and maintaining effective infection, prevention and control measures. This often required patients who tested positive for COVID-19 to be cared for together, in isolation, to allow other areas of the hospital to continue to operate and minimise hospital-acquired COVID-19 infection. This meant clustering patients with COVID-19 who required oxygen therapy in specific areas of the hospital. In addition, this allowed efficiencies in staffing and ensured only specific staff groups may be asked to care for patients with COVID-19. HSIB has commented on the need for such measures to enhance infection, prevention, and control practices when responding to COVID-19 (Healthcare Safety Investigation Branch, 2020).

The national investigation

4.1.11 The ability to clearly articulate, understand, and respond to competing demands across multiple professional specialities was important in ensuring a safe and effective response to any issues impacting on the MGPS. Where effective multidisciplinary team (MDT) working was in place these competing demands could be considered and a risk-based approach taken to planning patient care. However, where these competing demands were not well understood by the MDT there were opportunities for misunderstanding about how the MGPS may operate and how best to plan for the response to COVID-19.

4.1.12 The MDT forum for shared understanding and decision making about MGPS in acute trusts is normally the medical gas committee (MGC). The investigation identified that some trusts had stepped up the frequency of MGC meetings and engagement to ensure a proactive response to emerging MGPS issues.

4.1.13 In addition, the investigation identified other trusts where additional ad hoc MDT working arrangements were put in place during the COVID-19 pandemic to respond to emerging patient safety risks from MGPS. This included the establishment of specific oxygen teams within some NHS trusts, and at the national level within NHSE/I, to encourage responsive MDT working.

4.1.14 However, other trusts had still struggled to effectively utilise the MGC in the response to COVID-19 or adopt other mechanisms to ensure effective MDT working on MGPS issues. The investigation was told of challenges in clinical engagement in MGCs across various trusts. The investigation identified a range of factors that could impact on clinical engagement, including:

• competing clinical duties scheduled at times of MGC meetings, such as patient clinics
• competing non-clinical duties scheduled at times of MGC meetings, such as management meetings
• a lack of awareness of the MGC and its function, such as the perception that this was an ‘estates’ meeting
• challenges in engagement between clinical and non-clinical teams.

4.1.15 The investigation heard examples of how different MDT representatives were often ‘speaking different languages’ at the MGC. While estates colleagues were often focused on technical aspects of the MGPS (such as flow rates, pressure and infrastructure concerns) clinical colleagues were focused on clinical issues (such as the use of equipment to deliver oxygen therapy, the quality of oxygen gas and patient saturation levels). This had the potential to alienate different groups and sometimes led to the MGC becoming primarily an estates-focused forum.

4.1.16 NHSE/I has supported the important role that chief pharmacists should play in any future direction of MGC and MDT decision making, given their role in managing medicines, including oxygen (NHS Specialist Pharmacy Service, 2021b). Such a role may include acting as the interface between the MDT members to try and ensure engagement across the MGC and encourage shared understanding and ownership of MGC issues.

4.1.17 The investigation also found a lack of prioritisation of MGC meetings due to the previous staff and organisational experience of the MGPS not posing challenges to patient care. The investigation was told by trusts of occasions when MGC meetings were cancelled at short notice owing to other, seemingly higher priority, operational demands. In one case, a trust had completed only 2 of a scheduled 12 MGC meetings in the previous 3 years.

4.1.18 This meant that, for some trusts, there was an inability to easily step up the MGC process when required during the COVID-19 pandemic in the light of increased challenges to MGPS. This would have been hampered further in trusts where there was a lack of designated officers (see Designated officers, 4.2.21) to help integrate clinical and non-clinical decision making.

4.1.19 HSIB has published interim bulletins focusing on:

• the importance of multidisciplinary working for MGPS (see appendix A)
• the role of the MGC as a forum for encouraging a shared understanding of MGPS and helping in the response to MGPS issues (see appendix B).

4.1.20 The second interim bulletin included the following recommendation to support a clearer understanding of the role, function, and key attendees of the MGC in order to promote enhanced MDT working:

Authorised persons

4.2.1 The health technical memorandum (HTM) identifies the authorised person as ‘the individual who will take on the responsibility for the day-to-day management of the MGPS’ and identifies that ‘the safe operation of a medical gas pipeline system relies on skilled staff who understand the system and who can liaise with clinical users to ensure continuing patient safety’ (Department of Health, 2006). Authorised persons take on a key operational role in managing the MGPS and act as a liaison point between estates and facilities teams and the wider multidisciplinary team.

4.2.2 Authorised persons are appointed by the trust executive. However, they are required to be approved for the role and recommended to the executive by the authorising engineer. Authorised persons then require specific training and reassessment every three years in line with the HTM to maintain their skills.

The reference event

4.2.3 The Trust’s previous two annual MGPS reviews identified that there were not enough authorised persons available within the Trust to manage the MGPS. At one site, the Trust was short of one authorised person with another soon to retire, creating another vacancy. During the pandemic, the Trust also suffered additional shortages in authorised person staffing due to illness.

4.2.4 The Trust told the investigation that it was challenging to recruit and retain staff with the appropriate knowledge and experience to act as authorised persons. The Trust highlighted a lack of available engineering experience within the NHS and the impact of an ageing estates and engineering workforce.

The national investigation

4.2.5 Concerns about the availability of engineering expertise within NHS organisations was expressed by several organisations and individuals that engaged with the investigation. The investigation was told about an ageing engineering and estates workforce and concerns about the ability to ensure that suitably qualified and experienced staff would be available to step into the place of retirees. Trusts pointed to the lack of dedicated graduate programmes or vocational entry and development programmes aimed specifically at engineering and estates professions within the NHS.

4.2.6 The investigation identified several trusts that had been required to appoint engineers from other industries to address this gap. This created some benefits for the NHS in accessing external knowledge and experience to complement NHS services. However, new staff often needed time to become familiar with NHS services, operational processes and MGPS infrastructure.

Authorising engineer

4.2.7 The HTM sets out that an authorising engineer should be either:

• ‘a chartered engineer in an appropriate engineering discipline; or
• have sufficient engineering and pharmaceutical knowledge and be qualified to the level equivalent to incorporated engineer [a recognised level of professional registration for engineers].’ (Department of Health, 2006)

4.2.8 The authorising engineer plays a key role in assuring the MGPS, recommending authorised persons for appointment, and providing training to ensure other functions under the HTM are properly carried out. The authorising engineer is the only external assurance mechanism included in the HTM for MGPS.

4.2.9 Under the HTM, authorising engineer training should cover the ability to ‘conduct an annual audit and review of the management systems of the MGPS’ and ‘ensure that an annual review takes place’ (Department of Health, 2006). The investigation found it was common practice for an annual report to be requested to help trusts to assure their MGPS.

The reference event

4.2.10 The Trust had received annual reports from the authorising engineer in the years prior to the incident. These including ratings to assist the Trust in identifying where the MGPS required improvements and recommendations to help the Trust address these concerns.

4.2.11 The investigation compared the previous two annual reports. This identified a number of recommendations that had been carried over because they had not been actioned, including nominating designated officers (see Designated officers, 4.2.21), recruiting additional authorised persons (see Authorised persons, 4.2.1), and out-of-hours cover arrangements for authorised persons.

The national investigation

4.2.12 Other trusts told the investigation about circumstances in which it was challenging for trusts to comply with authorising engineer annual report. This was often due to factors highlighted in this report (see Designated officers, 4.2.21, Training for NHS staff, 4.2.34, MGPS infrastructure, 4.3). In such circumstances, authorising engineers have no regulatory or enforcement function under the HTM. The investigation was told that authorising engineers are also frequently private providers that must tender to trusts for their work. This may create challenges in authorising engineers challenging organisations or seeking assurance that issues highlighted in their reviews are actioned if they fear this may jeopardise their ability to secure further work with the organisation.

4.2.13 The investigation was also told, by trusts and other stakeholders, of concerns about the ability to ensure that authorising engineers were suitably qualified and experienced to perform their role. The concept of suitably qualified and experienced personnel (SQEP) originated within the UK nuclear industry and is also used within the defence industry, but similar requirements arise in other industry sectors and other countries. Typically, to be considered a SQEP a person needs a professional qualification and several years of relevant experience within their designated role.

4.2.14 The investigation identified some examples of very positive, proactive relationships between trusts and authorising engineers who appeared to be SQEP. This included the authorising engineer in the reference event. Trusts reported positive engagement with these authorising engineers and were assured that they helped maintain the safety of the MGPS. However, these authorising engineers were often in high demand and may be undertaking the role for several organisations. This created practical challenges in day-to-day engagement with authorising engineers, particularly during the COVID-19 pandemic when MGPS concerns were widespread.

4.2.15 The investigation was told of other examples where trusts had identified issues with authorising engineers or where authorising engineers may subcontract to other providers to carry out work. In such circumstances, trusts were unable to assure themselves that authorising engineer, or their subcontractors, were SQEP and could carry out the role. This included examples of trusts being unable to obtain annual reviews from their authorising engineers due to the authorising engineers’ capacity or capability to provide the report.

4.2.16 Under the HTM, authorising engineer training records ‘will be kept by the Institute of Healthcare Engineering and Estate Management (IHEEM)’ (Department of Health, 2006). This is the only reference to IHEEM within the HTM. IHEEM is an independent professional engineering institute for healthcare. In addition, IHEEM keeps a register of authorising engineers that it assesses as SQEP to perform the role.

4.2.17 The investigation found that it was not mandatory for authorising engineers to be registered with IHEEM, or any other body, before providing authorising engineer services. Although any authorising engineers who hold chartered engineer status will be registered with other engineering bodies, any authorising engineer not holding chartered status may not be. In addition, the investigation found that authorising engineers not registered with IHEEM did not routinely submit their training records to IHEEM in line with the HTM. This means that there is currently no independent, external oversight of the training authorising engineers have undertaken to ensure they are suitable for their role.

4.2.18 The authorising engineer is appointed by the trust executive. The investigation has highlighted the challenges in ensuring suitable engineering experience is available within the NHS (see Authorised persons, 4.2.1) and was told that many director of estates posts, or similar, are no longer board-level roles and instead may report into other board-level executives.

4.2.19 In such circumstances, the executives appointing authorising engineers may face challenges in identifying whether they are SQEP, if the authorising engineer is not a chartered engineer and/or has not been subject to independent scrutiny, such as via IHEEM registration.

4.2.20 The HTM sets out the authorising engineer as vital in overseeing the safe and effective operation of MGPS and providing external assurance. However, the factors highlighted by the investigation bring into question the current limitations in trusts assuring themselves that authorising engineers can effectively fulfil this role. The investigation also highlights the lack of proactive engagement in an optional registration process and the need for a proactive approach going forward.

Designated officers

4.2.21 The HTM sets out the role of designated officers for the MGPS. Designated officers include the designated nursing officer (DNO) and designated medical officer (DMO), and are defined in the HTM as ‘the person in each department with whom the Authorised Person (MGPS) liaises on any matters affecting the MGPS and who would give permission for a planned interruption to the supply’.

4.2.22 The HTM identifies that there should be a designated officer for every department within the trust. However, there may be a difference in the role of the DNO and DMO. A DNO may be given responsibility for a whole hospital, multiple wards or just a single ward. A DMO may be involved in the gas supply to a whole department, although some hospitals may operate without a DMO if a DNO is in place.

4.2.23 The designated officer plays a key role in providing permissions for works to take place which may disrupt the MGPS within specific ward or departmental areas within a trust. Designated officers also take on a key role in emergency management, such as identifying any MGPS issues and isolating the relevant supply.

4.2.24 In addition to the designated officer’s role in authorising works that affect the MGPS, the HTM identifies that ‘it is essential that there is liaison between the medical and nursing staff that use the MGPS and the Authorised Person to ensure that the MGPS is appropriate to their needs’ (Department of Health, 2006). The designated officer acts as the focal point for these communications between estates and clinical teams to advise on any special requirements for the MGPS relevant to their clinical area.

The reference event

4.2.25 The Trust told the investigation that it had always been challenging to appoint designated officers. The authorising engineer annual report in both 2019 and 2020 highlighted the need to nominate appropriate designated officers and for them to receive appropriate training. Approximately six years ago, the Trust had recruited and trained sufficient DNOs but following a directorate restructure and staff attrition, these post holders became unavailable to fulfil their role. The Trust did not have a DMO in place.

The national investigation

4.2.26 The investigation found that there was a lack of consistent use of designated officers in trusts. The investigation was told that whether designated officers were in place often depended on whether there were individual clinicians with a specific interest in the MGPS and a desire to be involved in this work.

4.2.27 Some trusts described having DNOs in place, while other trusts echoed the reference event Trust in explaining the challenges in recruiting and training staff to these posts. In addition, the investigation was told that DMOs “are like gold dust” and many trusts the investigation engaged with were not aware of their role.

4.2.28 Without specific individuals who were willing to engage in the MGPS, the investigation found that there was no systemic approach to ensuring that appropriate designated officers were nominated and trained across the NHS. This impacted on the ability to complete operational works on the MGPS and meant that there was a missed opportunity for clinical and non-clinical staff to work together and develop a shared understanding of the competing demands on the MGPS.

Oxygen champions

4.2.29 In 2008, the British Thoracic Society (BTS) issued guidance on emergency oxygen use in adult patients (O’Driscoll et al, 2008). To aid in the roll-out of the guidance local ‘oxygen champions’ were identified within acute trusts. The intention was that oxygen champions would ‘review local oxygen policy in accordance with the national guidelines, arrange staff education, and ensure oxygen prescription and monitoring could be achieved on drug and observation charts’ (Kane et al, 2013).

4.2.30 In 2017, the BTS updated its guidance for oxygen use for adults in healthcare and emergency settings (British Thoracic Society, 2017). An appendix to the guidance identifies that ‘all acute NHS Trusts should have one or more local oxygen champions, ideally a medical oxygen champion and a nursing and/or AHP [allied health professional] champion’.

The national investigation

4.2.31 The investigation found that there may have been some scope for confusion in the role and function of oxygen champions during the COVID-19 pandemic. Some trusts told the investigation that they had oxygen champions in place. However, in some instances it was unclear whether trusts understood that oxygen champions, as described by the BTS guidance, were limited to a clinical role and that their purpose was not to act as a link between clinical and estates teams for the MGPS.

4.2.32 The BTS told the investigation that it “does not provide guidance or comment on oxygen supplies within hospitals, or on matters relating to engineering or the organisation of medical gas committees. Our advice and guidance is aimed at health care professionals and the clinical care they provide to patients”.

4.2.33 The current role of oxygen champions does not include acting as an interface with estates or engineering colleagues. Currently this liaison role should be fulfilled by the designated officers under the HTM. However, there may be opportunities to incorporate the expertise of BTS oxygen champions, and other specialist groups, into the operation of MGPS to enhance the interface between clinical and estates/engineering teams.

Training and competence for NHS staff on MGPS

4.2.34 The HTM sets out the training requirements for different roles assigned by the HTM in the management of MGPS. This includes designated training for authorised persons and designated officers. The HTM also sets out the training required for other NHS staff operating the MGPS.

4.2.35 In one section, the HTM identifies that ‘medical and nursing staff who will use the MGPS should be trained in the practical use of the system (including the safe use of basic equipment such as flow meters and suction controllers), safety procedures and actions in the event of emergencies’. However, in another section of the HTM more extensive training is set out ‘for all nursing/medical staff working with medical gas systems’ (Department of Health, 2006). This includes training on eight separate areas related to the MGPS.

The reference event

4.2.36 The Trust told the investigation that nursing and medical staff would not routinely be trained to understand the basics of the MGPS and that it was challenging to access training. This was due to a combination of financial restraints and the availability of staff to complete training.

4.2.37 Trust staff reflected that prior to the COVID-19 pandemic, the majority of them had never had to question how oxygen was provided to clinical areas or consider how knowledge of the process could aid in providing clinical care; when they needed piped oxygen, it had always been available without concern. However, this meant that the challenges faced by the MGPS in responding to the demands of COVID-19, and the potential impact on clinical care, were not well understood by all clinical staff.

The national investigation

4.2.38 The investigation found that the training requirements set out in the HTM were not in place within the NHS. The NHS electronic staff record includes a range of mandatory and optional training modules for NHS staff to complete. Medical gases is a specific area included in the course catalogue, but no training modules are available.

4.2.39 Prior to COVID-19, the lack of training on MGPS delivered to all staff may not have been a concern, given the general effective operation of MGPS in this period. Similarly, in a post-COVID-19 environment it would appear unreasonable to ask that all staff have in-depth knowledge of the MGPS if this does not routinely directly impact on their specific area of work.

4.2.40 Identifying what level of understanding and competence around MGPS is required by different staff groups could facilitate education based on differing knowledge requirements within different clinical specialties. It could also facilitate an understanding of how the operation of MGPS can directly impact on patient care.

Investment in MGPS

The reference event

4.3.1 The Trust had a significant, multimillion-pound estate maintenance backlog. Staff felt that this impacted on the finances and resources available to address estates-related risks and limited the Trust’s ability to undertake a proactive programme of estates works.

4.3.2 An example of this was provided by the high dependency unit (HDU) in the reference event. Although this was a ‘new’ HDU, no additional funding had been available to increase the capacity of the MGPS pipework, which was designed to deliver the HTM minimum flow rates. The HDU had been built at the end of an existing Nightingale ward (a type of hospital ward that comprises one large room) and was reliant on the existing pipework infrastructure, including the MGPS providing oxygen to ward bed spaces before it made its way to the HDU. This meant that any redundancy in the system could be drawn off before the MGPS reached the HDU.

The national investigation

4.3.3 In the context of the wider NHS, data from NHS Digital (2020a) identified that the NHS estates backlog maintenance costs increased by 40% between 2018/19 and 2019/20, from £6.46bn to £9bn.

4.3.4 Trusts told the investigation that they thought there was, rightly, a focus on investing estates funds into new clinical areas to deliver enhanced patient care and experience. However, due to funding limitations this sometimes meant that new environments were reliant on older infrastructure and systems that were unable to attract the same prioritisation for capital investment.

4.3.5 Other trusts and NHS estates professionals reported similar concerns to the investigation. One estates team told the investigation that in the NHS “we [often] do the pretty and the shiny … paint walls, build new units”. Vacuum insulated evaporator (VIE) plant manufacturers told the investigation that there were still some “unbelievably old” pieces of equipment in use in the NHS. This sentiment was shared by other estates teams who described that there was sometimes a failure to invest in underlying infrastructure, which meant that existing infrastructure was often extended beyond its anticipated lifespan.

4.3.6 During the COVID-19 pandemic, NHS England and NHS Improvement (NHSE/I) received data to show the capacity at which trusts were operating their MGPS. Where the capacity of the VIE was identified as a concern, and trusts were unable to generate sufficient oxygen to meet demand, NHSE/I has engaged in a programme of capital works with trusts to replace elements of the VIE plant. This has enabled some trusts to increase their ability to generate oxygen. However, the ability to deliver volumes of oxygen into a hospital is still limited by other factors. Increasing the amount of oxygen available cannot overcome the specific challenge related to limitations in pipework infrastructure within a hospital estate.

4.3.7 HSIB has previously commented on the role of a safety management system (SMS) in providing a systematic and proactive approach for identifying and managing safety risks (Healthcare Safety Investigation Branch, 2020). An SMS integrates system safety, human factors and financial management within the overarching organisational framework. They are commonly used in other safety-critical industries to recognise where safety may be compromised, and where there are threats to operational goals.

4.3.8 In 2013, the Health Foundation (Vincent et al, 2013) released a report to summarise how safety is monitored and managed in healthcare. The report highlighted concerns that although healthcare may have embedded processes to focus on quality, it has not developed a safety function to fully consider how a system may fail to function.

4.3.9 The investigation acknowledges that prior to the COVID-19 pandemic MGPS had largely operated effectively without concern. However, the COVID-19 pandemic placed unprecedented demands on these systems, with many challenges caused by the design of MGPS and a lack of shared understanding of how they operate within NHS trusts.

4.3.10 An MGPS is an example of a critical system within acute hospitals where there are significant consequences should it fail. Elements of an SMS have operated locally to identify the risk, such as placing MGPS risks on the trust risk register for action. However, no integrated SMS is operated nationally that would have aided in the proactive identification and response to this risk pre-pandemic across the NHS. Due to a range of factors the investigation found that MGPS have often been overlooked for investment and upgrade. In addition, key MGPS processes (such as the MGC) may not have been given appropriate priority to allow for multidisciplinary engagement to assist in the proactive monitoring of oxygen usage and safety.

4.3.11 Prior to the COVID-19 pandemic there was a need for many NHS organisations to balance priorities in the management of their MGPS with other operational demands, because of the need for investment in other areas or day-to-day prioritisation of other estates or clinical work. This may often have been necessary to meet the range of competing goals facing many NHS trusts and did not raise concern when MGPS operated effectively.

4.3.12 However, a lack of ongoing ability to identify, manage and respond to MGPS concerns left many NHS organisations vulnerable to issues arising from MGPS when they emerged during the COVID-19 pandemic. This was particularly important during the pandemic owing to the critical importance of oxygen as the primary treatment option available for patients.

4.3.13 The ability and resource for NHS trusts to adopt a greater focus on MGPS management would assist in recognising the importance of these systems and identifying potential risks to patient safety and operational performance should they fail. Such an approach at a systems level would have helped the NHS to better identify and respond to these challenges prior to the COVID-19 pandemic.

4.3.14 The issues identified in this investigation regarding MGPS support the adoption of a more proactive approach to system safety, such as the integration of an SMS in healthcare to identity and mitigate against key risks to patient safety before they arise.

Radial and ring main MGPS

The reference event

4.3.15 The Trust told the investigation that it had a radial MGPS (see 1.2.4) in both of its primary hospital sites. The HTM (Department of Health, 2006) does not specify that a ring main system is required in new MGPS works and envisages that most works to establish MGPS will involve the building of radial systems.

The national investigation

4.3.16 The investigation found that radial systems were common in the NHS and may represent many systems in use within acute hospital trusts. Prior to the COVID-19 pandemic, the investigation could not find any evidence that the use of radial systems posed any significant challenges to NHS trusts. This was due to the previous, anticipated, levels of oxygen required by NHS trusts from their MGPS being achievable with radial systems.

4.3.17 The COVID-19 pandemic placed unprecedented demand on MGPS. In particular, the increased levels of COVID-19 and non-COVID-19 activity that NHS trusts saw in wave two placed significant strains on MGPS. In these instances, the role of MGPS in delivering appropriate care was highlighted as their capacity to deliver adequate oxygen supplies became stretched.

4.3.18 The investigation identified a number of examples where trusts with radial MGPS had taken mitigating engineering actions to try and increase the capability of their oxygen systems in light of COVID-19. These included:

• Manifold systems: where banks of oxygen cylinders are attached to the MGPS to provide a back-up level of oxygen should pressure levels drop.
• Use of pipeline flow meters: where flow rates of oxygen into and throughout the hospital are measured in real time to help identify areas of concern and allow action to be taken to maintain balance in the MGPS.
• Enhanced VIE capacity: NHSE/I has undertaken a national programme of works helping trusts to replace elements of their VIE plant, enabling a greater volume of liquid oxygen to be stored.
• Enhanced evaporator capacity: to allow more gas to be created from the liquid stored in the VIE.
• Ring main systems: trusts seeking to move to ring main systems (see 1.2.4) to provide greater resilience.

4.3.19 The HTM was written in 2006 and, due to the need to prioritise updates to other HTM’s and the impact of the COVID-19 pandemic, has not been able to be updated since it transferred to NHSE/I control. A review of the HTM would help ensure that knowledge about MGPS installation and operation is updated to account for developing knowledge since 2006. It would also enable the inclusion of learning that can be applied from the COVID-19 pandemic to assist in the response to any future peaks of COVID-19 or other respiratory illnesses. This would also help to identify and address examples such as those highlighted in this report where current HTM guidance does not reflect practice at trust level and where changes to the HTM, or current practice, may be required to ensure improved safety of MGPS.

NHS Premises Assurance Model

4.4.1 The NHS Premises Assurance Model (PAM) incorporates specific questions about MGPS to ask if trusts have:

• policies and procedures in place
• appropriately qualified, competent, and appointed people
• a risk assessment
• adequate maintenance processes
• up-to-date training and development
• business continuity plans
• robust annual assessment
• costed action plans.

4.4.2 The NHS PAM asks trusts to assess themselves against these questions, which gives them a rating on a five-point scale from inadequate to outstanding. To accompany the rating, trusts are asked to provide evidence of how they meet the requirements set out in the PAM.

The reference event

4.4.3 The Trust had completed the NHS PAM and a member of the Trust was involved in helping to further develop the PAM model via involvement in national working groups.

The national investigation

4.4.4 NHSE/I told the investigation that up until April 2020, the NHS PAM was a voluntary process. NHSE/I estimated that around 30% of trusts completed the PAM and there was no requirement to share PAM data. Due to this, local ownership of the completed PAM was important, ‘otherwise no action or advantage will be obtained from using it’ (NHS England and NHS Improvement, n.d.).

4.4.5 From April 2020 onward, NHS trusts have been required to complete the NHS PAM via a spreadsheet that can be downloaded from the NHSE/I website (NHS England and NHS Improvement, n.d.). However, there is currently no mechanism for paper-based assessments to be shared with NHSE/I and no national oversight currently exists of the PAM scoring.

4.4.6 Centralised online reporting will allow NHSE/I to collate national data on estates-related issues, including MGPS issues. This data can then be compared to other existing datasets (NHS Digital, 2020a; 2020b) to provide a more informed view of the potential estates-related patient safety issues across various NHS trusts. This information could then be shared with appropriate national bodies to assist in the identification and response to emerging estates-related patient safety concerns. This would represent a further element of a national approach to an SMS in healthcare.

4.4.7 Given the critical importance of MGPS in the response to COVID-19, and the potential for significant patient safety concerns when MGPS fail, a proactive approach to identifying MGPS issues should be supported.

Care Quality Commission assessment

4.4.8 The quality and safety of estates and MGPS also falls within the remit of the Care Quality Commission (CQC) under the relevant regulations governing its regulatory activity (Health and Social Care Act 2008 (Regulated Activities) Regulations 2014).

4.4.9 The CQC key lines of enquiry for NHS trusts (the questions that form the basis of its inspections) include questions on the design, use and maintenance of healthcare facilities (Care Quality Commission, 2018a). They include reference to medical gases to consider how medical gases may be safely ordered, transported, stored, and disposed of. In addition, oxygen would be included in key lines of enquiry about the appropriate prescribing, administration, and supply of medicines to patients.

4.4.10 The CQC told the investigation that the approach to inspection has typically been for these issues to be considered from the perspective of the immediate care environment, such as considering whether ward areas were clean and well maintained. However, it has not been typical for CQC inspections to provide a broader assurance about the management of estates and facilities at the trust level and no specific focus would routinely be placed on the management of MGPS.

4.4.11 The CQC echoed comments from other stakeholders in explaining that MGPS had never been a cause for concern prior to the COVID-19 pandemic. In addition, as a care regulator the CQC did not have specific estates and facilities expertise that would routinely be able to assist in the consideration of associated issues at trust level.

4.4.12 NHSE/I told the investigation that the NHS PAM data will routinely be available to the CQC once the PAM online portal is established. This data will be helpful in providing additional insight to the CQC about any specific challenges relating to MGPS, and estates and facilities issues more broadly. This would allow the CQC to consider this data, alongside other available information, to identify where specific focus may need to be placed on estates issues during planned inspection activity.

4.4.13 However, as set out above, the CQC does not currently employ estates and facilities experts who would be able to inform the CQC about potential estates and facilities concerns arising from the datasets. The CQC would also be unable to routinely call on estates and facilities expertise to assist in any targeted or planned inspection activity.

4.4.14 The CQC well-led inspection framework acknowledges that the CQC is similarly limited in its ability assess NHS trusts’ use of financial resources within its well-led framework. To ensure this is appropriately accounted for, the CQC well-led review incorporates specialist support from financial experts within NHSE/I to assist the CQC in its consideration of financial issues.

4.4.15 Given the lack of engineering expertise sometimes available to NHS trusts, existing expertise in NHSE/I and industry could be utilised in a similar way to the approach to assessing financial resources to assist the CQC to access appropriate, specialist estates and facilities knowledge when required. It would be for the CQC to determine when this expertise may be required based on data provided via the NHS PAM, or other CQC datasets that may highlight estates related concerns.

NHSE/I oxygen guidance

The reference event

4.5.1 The Trust explained that the NHSE/I guidance (NHS England and NHS Improvement, 2020d) was not issued via the Central Alerting System (CAS). Instead, it was provided to the Trust directly by its authorising engineer. However, this guidance was not identified by the Trust, as the authorised persons who received the email were both absent from the Trust. This created a delay in the Trust identifying the guidance and considering what action should be taken.

The national investigation

4.5.2 NHSE/I explained to the investigation that disseminating the oxygen guidance (NHS England and NHS Improvement, 2020d) as soon as possible was a priority, to try and ensure that NHS trusts could respond to the emerging risks to MGPS caused by COVID-19.

4.5.3 At the time, there were challenges relating to the volume of guidance being issued and the timescale in which guidance could be issued via centralised NHSE/I mechanisms (Healthcare Safety Investigation Branch, 2020). Because of this, the estates teams relied on a variety of other methods to disseminate this guidance, including existing online platforms that bring together NHS estates professionals and distribution to authorising engineers. This was in the hope that guidance would reach the relevant senior estates staff.

4.5.4 However, the volume of different guidance being received by NHS organisations during the first wave of the COVID-19 pandemic may have made it challenging for all relevant guidance to be identified and acted on. This was a particular challenge where trusts received guidance via a range of separate mechanisms (Healthcare Safety Investigation Branch, 2020). In addition, issues with the current estates workforce and availability of authorising engineers to support NHS organisations may also have created challenges in this guidance being applied in practice.

Central Alerting System alerts

The reference event

4.5.5 The Trust told the investigation that it had identified the CAS alerts via its governance teams. The Trust relied on a staff member to monitor the CAS and distribute alerts across the Trust according to what was identified as being required in the CAS system.

4.5.6 The period in which the alerts were issued in spring 2020 was very busy because of the COVID-19 pandemic and the Trust told the investigation this created challenges in identifying and responding to all the relevant alerts. In such instances, the Trust was required to prioritise its resources in an attempt to resolve the most serious threats to patient and staff safety.

The national investigation

4.5.7 Other trusts told the investigation that they had also faced challenges in ensuring that the appropriate CAS alerts were identified and acted on during this period. A search of the CAS identifies that in March and April 2020 when the relevant alerts were issued, there was a 177% increase in the number of alerts compared to the same period in 2019 (see figure 5).

4.5.8 These alerts were necessary to address the emerging risks from COVID-19 and ensure that the NHS was using updated evidence in its response to the pandemic. However, the additional volume of alerts, coupled with an increased volume of associated guidance in this period (Healthcare Safety Investigation Branch, 2020), meant it would have been a challenge to ensure that all relevant alerts were appropriately identified and responded to.

4.5.9 NHSE/I told the investigation that feedback from healthcare providers was that safety-critical communications did not always stand out from other communications, and the required actions were not always clear or effective. It was also found that it was not always made clear when an alert had significant clinical implications, such as a need to recall and review patients, risking those alerts being treated in the same way as more straightforward recalls of medicine batches or medical equipment.

4.5.10 The CQC (2018b) identified that NHS trusts did not always have strong systems for implementing CAS alerts. In response, the NHSE/I National Patient Safety Alerting Committee (NaPSAC) was established to improve the effectiveness of these safety-critical communications and to support providers to better implement the required actions through the introduction of national patient safety alerts.

4.5.11 Before an alerting body can issue a national patient safety alert it must first be accredited by NHSE/I to ensure their systems and processes for producing and developing alerts meet common standards. Presently, the NHSE/I national patient safety team, the Medicines and Healthcare products Regulatory Agency, and Public Health England are accredited. However, owing to the impact of COVID-19, other alerting bodies have yet to receive accreditation.

4.5.12 There has been a period in which some organisations can issue national patient safety alerts while others continue to issue using existing types of alerts, messages, and other communications via the CAS. The expectation is that all alerts must be actioned by providers accordingly.

4.5.13 Some trusts told the investigation that they considered that there was an informal hierarchy of CAS alerts. Trusts told the investigation that they perceived national patient safety alerts to be the highest priority of alert and that they understood that these alerts meant there was a risk to patients of serious injury or death.

4.5.14 A national patient safety alert was issued on 1 April 2020 (NHS England and NHS Improvement, 2020f) and was prioritised by the Trust for action. The alert identified four deaths in a two-year period due to problems in transferring patients requiring high flow nasal oxygen. There was a clear risk that individuals may be severely harmed or die from such event and the Trust deployed a multidisciplinary team to provide a response.

4.5.15 The estates and facilities alert issued on 31 March 2020 identified that issues with MGPS ‘could lead to a failure of oxygen delivery systems throughout the hospital … there is also a risk of rapid and unpredictable depletion of the VIE. Both of these situations present a potentially significant risk to multiple patients simultaneously’ (NHS England and NHS Improvement, 2020a). There was a clear risk that multiple patients could be harmed or die because of such MGPS issues and this alert was sent to a broader mailing list than would usually receive an estates alert of this type. However, this alert was not issued as a national patient safety alert and so was not prioritised by the Trust.

4.5.16 In addition, CAS alerts can be issued with target dates for completion of actions and a requirement to report compliance. This had been required for the April 2020 national patient safety alert. However, the estates and facilities alert was not issued with any requirement for a response to indicate that the actions had been completed. During this period, trusts were responding to multiple alerts and guidance in response to the first wave of COVID-19 (Healthcare Safety Investigation Branch, 2020). If an alert issued during this time did not require actions to be completed and reported to the CAS, this may have influenced the perception of the alert’s relative importance and the resource that was allocated to a response.

4.5.17 The investigation identified that all the CAS alerts relevant to this investigation were issued as estates and facilities alerts. Currently, NHSE/I’s estates and facilities teams are not accredited to issue national patient safety alerts via the CAS. Although some alerts specified that they would require action from executive trust staff, this would not have been common for estates and facilities alerts and would not have been immediately evident from the way in which these alerts were presented on the CAS.

4.5.18 The investigation also identified design factors that influenced the perception of national patient safety alerts within the CAS. National patient safety alerts normally require executive-level involvement (NHS England and NHS Improvement, 2019) and are highlighted in a different colour on the CAS display to indicate that there is a different process to follow in responding to these alerts. Figure 6 shows how the CAS displayed the alerts referred to above.

4.5.19 The International Organization for Standardization provides guidance on human-centred design for interactive systems (International Organization for Standardization, 2019). Human-centred design is an iterative process that relies on an explicit understanding of users, tasks and environments and is driven and refined by user-centred evaluation to ensure systems are safe and effective when used.

4.5.20 Colour coding and shading can present cues to people interacting with systems to indicate what is expected from them. Commonly, this will include increasing the visibility of critical information. International Organization for Standardization (ISO) standards include reference to colour coding for safety signs and safety markings (International Organization for Standardization, 2011). For example, in this guidance red is commonly used to indicate danger and yellow is commonly used to indicate caution.

4.5.21 The ISO standard specifies that blue indicates that an action is mandatory. In the case of the CAS, the additional blue shading may have an unintended consequence of reaffirming any belief held by users of the system that there is a hierarchy of CAS alerts, or that some alerts require more attention than others within the current system. In the current dual format, it may be possible for alerts of significant importance, such as those issued by NHSE/I estates and facilities in this instance, to be deemed less important owing to the design of the system.

4.5.22 The investigation acknowledges the challenges faced in updating the CAS to account for the move to the national patient safety alert reports and the limitations of the CAS as an older computer system. However, additional proactive involvement of user-centred design experts may assist in the development of further updates to the CAS to ensure the principles of user-centred design are considered.

• The MGPS is a critical system in the safe and effective operation of an acute NHS hospital.
• The COVID-19 pandemic placed unprecedented demands on MGPS due to the number of patients requiring oxygen therapy and the different types of oxygen therapy equipment used.
• There is a lack of shared ownership and knowledge of MGPS among hospital-based multidisciplinary teams; this limits trusts’ ability to effectively respond to MGPS patient safety concerns.
• Organisations that utilised a multidisciplinary approach to understanding and planning the MGPS involvement in the COVID-19 response, including the impact of ward moves and the choice of oxygen therapy, were better able to respond to demands on the MGPS system.
• Current guidance on the design and management of MGPS contained within the relevant health technical memorandum is outdated and does not reflect developments in oxygen therapy and challenges in managing MGPS.
• The investigation acknowledges work carried out by national bodies to help NHS trusts address the emerging risks to MGPS, including the issue of interim guidance in addition to the HTM and capital investment in MGPS.
• Current assurance mechanisms for MGPS are not effective in ensuring that MGPS-related patient safety concerns are proactively identified and resolved.
• A lack of financial investment in updating MGPS infrastructure created challenges for NHS trusts in responding to the COVID-19 pandemic.

In addition, HSIB has released two interim bulletins during the investigation highlighting key findings emerging from the initial reference event investigation and the role of medical gas committees: ‘Oxygen issues during the COVID-19 pandemic. January 2021’ and ‘Oxygen issues during the COVID-19 pandemic. March 2021’.