Investigation report: Clinical investigation booking systems failures - written communications in community languages

This investigation focuses on the systems used by healthcare providers to book patient appointments for clinical investigations, such as diagnostic tests and scans. ‘Clinical investigation booking systems’ are used throughout the NHS to support the delivery of patient care. Healthcare services use paper-based or fully electronic systems, or a combination of the two (hybrid systems), to communicate to patients the time, date and location of their appointment. These systems also produce information for patients about actions they need to take to prepare for their appointment. Written patient communication is a key output of clinical investigation booking systems.

This investigation examines the safety implications of patient communications, produced by booking systems, that do not account for the needs of the patient. In addition, it looks at why patents are ‘lost to follow-up’ after an appointment is cancelled, rescheduled or not attended. Lost to follow-up is the term used to describe a patient who does not return for planned appointments (whether for continued care or evaluations) or is no longer being tracked in the healthcare system when they should be.

To explore the issues involved, the investigation used the following real patient safety incident, referred to as ‘the reference event’.

A child of Romanian ethnicity was referred for an MRI scan, which required a general anaesthetic. The scan was booked, and a letter was sent to the child’s parents including the appointment details and pre-appointment instructions.

The Trust’s booking system was only able to produce appointment letters in English, and there were no Trust processes or policies to routinely translate written appointment information.

The family recognised key details in the written information, including the time, date and location of the scan. However, they were not able to understand the instructions about the child not eating or drinking (fasting) for a certain amount of time before the scan.

When the family attended the appointment for the scan, the child had eaten. This meant the MRI scan could not be completed and it was therefore cancelled. The radiology booking team did not receive confirmation of the need to rebook the scan, and 11 weeks passed before it was identified that the scan had not taken place. When the scan was rebooked, a letter was again sent to the family with the appointment details and information about the need for fasting. The letter was sent in English. The child arrived at the scheduled scan appointment, but had again eaten prior to arrival. The scan was cancelled and rebooked for the following day, when it was carried out and cancer was diagnosed.

The child received treatment, but sadly the disease progressed and they were placed on a palliative care pathway and died. There was no suggestion that identifying the cancer at an earlier point would have affected the child’s outcome.

The patient safety risk related to clinical investigation booking systems was identified following a routine review of incidents reported to the Strategic Executive Information System (StEIS). The StEIS is a national database for reporting serious safety incidents in healthcare.

The HSIB investigation gathered information about the systems and processes relevant to the reference event, and assessed the safety risk against its investigation criteria. After the initial information had been gathered and evaluated, HSIB’s Chief Investigator authorised a national safety investigation. Further details of the investigation approach are given in Appendix A.

The national investigation focused on:

• the safety implications of producing patient communications solely in English
• the context of and contributory factors to patients being lost to follow-up due to administrative functions.

• Written communications to patients about radiology appointments are routinely sent in English only.
• Healthcare staff expect that written appointment information will be translated by a patient’s friend or family member.
• NHS England standards do not require written appointment information to be given in any non-English language, other than for people with a disability.
• Confusion about the requirements for appointments can result in delayed care and additional costs if appointments need to be rebooked.
• The language needs of patients are not always clearly understood. There is differing national guidance as to whether written communication needs should be recorded.
• The national NHS system that holds patient information and populates it into many trust systems (the Personal Demographics Service) can store information on patients’ preferred written communication methods. However, this information is often not entered into the system.
• Administrative staff are not routinely involved in assessing and testing electronic booking systems before they are implemented.

1.1.1 The latest census found that, in 2021, 7.1% of the population in England and Wales (4.1 million people) were proficient in English but did not speak it as their main language. An additional 1.5% (880,000 people) could not speak English well and 0.3% (161,000 people) could not speak English at all (Office for National Statistics, 2022).

1.1.2 The investigation was unable to find similar information for reading English. However, in 2019, the Organisation for Economic Co-operation and Development reported that 16.4% of adults in England had low literacy skills (Organisation for Economic Co-operation and Development, 2019).

1.1.3 Community languages are those used by minority groups or communities where a majority language exists, such as English in the UK.

1.2.1 The subject of providing non-English language interpretation or translation (for reasons other than disability) is outside the scope of the NHS Accessible Information standard (NHS England, 2017). However, one of the guiding principles of the NHS, as reflected in its constitution, states that: ‘The patient will be at the heart of everything the NHS does.’ As part of this, the constitution says that NHS services ‘must reflect, and should be coordinated around and tailored to, the needs and preferences of patients, their families and their carers’ (Department of Health and Social Care, 2021b).

1.2.2 National Institute for Health and Care Excellence (2021) guidance on communications for patients with cancer states that appropriate information should be provided ‘in terms of language, ability and culture’.

1.3.1 Radiology is the branch of medicine that uses internal imaging of the body to diagnose and treat medical conditions. It encompasses various techniques such as X-ray, CT and MRI scanning.

1.3.2 Radiologists are doctors who specialise in interpreting images, so as to diagnose and manage medical conditions. A radiographer is the person who performs the scan, although some radiographers also interpret scans. Radiography is a regulated profession that requires a qualifying degree.

1.3.3 Radiology services are used by many medical specialities to provide information and guide treatment. Radiologists can also provide treatment themselves in the form of interventional radiology (minimally invasive, image-guided medical treatments).

1.3.4 Demand for radiology services in the UK has increased due to a growing and ageing population (The Royal College of Radiologists, 2019) and is expected to continue to increase. This, coupled with a shortage of radiologists, interventional radiologists and radiographers, means that there are often delays in imaging patients. Many hospitals outsource scans to the private sector and rely on doctors working overtime to keep up with demand.

1.3.5 Imaging delays can increase the time to diagnosis and treatment, with a resulting potential impact on patient outcomes.

1.4.1 As well as the demand on radiology services, an additional cause of imaging delays is patients being lost within the healthcare system (‘lost to follow-up’).

1.4.2 Lost to follow-up describes a patient who does not return for planned appointments (whether for continued care or evaluations) or is no longer being tracked in the healthcare system when they should be (Healthcare Safety Investigation Branch, 2021).

1.4.3 Patients being lost to follow-up is a risk in many parts of the healthcare system, and has been the subject of a previous HSIB investigation (Healthcare Safety Investigation Branch, 2021).

1.4.4 From 1 March 2021 to 28 February 2022 (inclusive), 34 incidents of treatment delay reported on the Strategic Executive Information System (StEIS) (the NHS tool that is used to report serious safety incidents) included the phrase ‘lost to follow’. In each of these 34 incidents, the delay related to the tracking of patients, rather than a lack of capacity to undertake clinical investigations.

1.4.5 This investigation relates to patients who are lost to follow-up after an imaging appointment has been booked, but is then not attended by the patient or needs to be rescheduled, as occurred in the reference event.

1.5.1 In order to manage appointments, healthcare providers use different systems to track patients and allocate the resources needed for clinical investigations (such as time slots and imaging machines).

1.5.2 Such systems can be electronic, paper-based or a combination of both, known as ‘hybrid’ systems.

1.5.3 Bookings for clinical investigations are managed by a team of administrative staff who coordinate the bookings, staff and machinery required, and liaise with patients and staff about planned appointments.

2.1.1 Following clinical assessment, a 3-year-old child of Romanian ethnicity was discussed at an oncology multidisciplinary team meeting. It was decided that an urgent MRI scan under anaesthetic was required to “rule out” cancer. The suspicious area was thought unlikely to be cancerous, and so responsibility for the child’s care remained with the consultant who had taken the case to the multidisciplinary team for discussion and did not transfer to the oncology team.

2.1.2 A paper referral form was completed by the child’s consultant and submitted to the radiology department to book the scan. The scan was booked to take place 10 days later and a confirmation letter, together with instructions that the child needed to be fasted for the scan (that is, not fed for 6 hours beforehand), were sent to the parents. These documents were written in English.

2.1.3 Before the scan date, the pre-admission and radiology teams attempted to contact the parents by telephone and email to ensure they understood that the child must not be fed before the scan. No reply was received to these communications, which were made in English.

2.1.4 The child arrived at the planned MRI appointment with their parents. After being booked onto the ward for a general anaesthetic, it was identified that the child had not been fasted as required. Therefore, the scan was cancelled by the ward staff and notice of the cancellation was communicated to the radiology booking team by telephone.

2.1.5 The radiology booking team telephoned the medical secretary of the child’s consultant and explained about the cancellation. The booking team required confirmation that the scan remained urgent before it could be rebooked and was asked by the medical secretary to send an email to them requesting this confirmation so that the child’s case could be discussed with the consultant and a response provided. An email was sent with this request.

2.1.6 Two weeks later a secretary for the oncology team identified that the scan had not been rebooked and contacted the radiology booking team to inform them this was outstanding, but a response was not received, and the appointment was therefore not rebooked.

2.1.7 Nine weeks later (11 weeks after the cancellation), a response was received to the original email sent to medical secretary of the child’s consultant. This confirmed that the scan was urgent and resulted in an MRI scan being booked to take place 2 weeks later. A standard letter, in English, was produced by the radiology booking system and sent to the child’s parents, informing them of the appointment and the need for the child to be fasted for the scan.

2.1.8 The pre-admission and radiology teams again attempted to contact the parents by telephone and email, in English. Calls to the parents were not answered and messages were left in relation to the forthcoming scan appointment.

2.1.9 The child arrived for the scheduled scan appointment, but had not been fasted. The scan was cancelled and rebooked for the following day, when it was carried out.

2.1.10 The results of the scan confirmed the presence of cancer. The child received treatment, but sadly the disease progressed and they were placed on a palliative care pathway before they died. There was no suggestion that identifying the cancer at an earlier point would have affected the child’s outcome.

3.1.1 When the paper MRI request form was completed, it was handwritten on it that an interpreter was required as the family’s first language was Romanian. Oncology staff explained that when a patient or their family did not have a good understanding of English, this was recorded in the patient’s paper notes and secretarial staff would arrange for an interpreter to be present at clinic appointments.

3.1.2 When the appointment for the child’s MRI scan was made by the radiology booking team, a standard letter was produced by the booking system. This gave details of the appointment, together with instructions relevant to the planned scan. This letter was written in English. Booking staff explained that while it may be noted on the system that a patient has another language requirement, there is no facility within the system to have appointment information sent out in another language.

3.1.3 The Chief Clinical Information Officer of the Trust told the investigation that it was common for appointment letters – not just those from the radiology department – to be sent in English only. They explained that translating appointment letters into the language required by each patient would not be practical because of the large number of languages spoken within the Trust’s local area.

3.1.4 The investigation was told that the pre-admission telephone calls from the pre-admission nursing team and appointment reminder calls from the radiology booking office staff were all made in English. In the case of the reference event, staff were unable to reach the family and messages were left, in English, about the forthcoming appointment. There was no process in place to ensure that such communications were made in the language required by patients or their families. Additionally, there was no process to follow up with patients if these calls were not answered or if messages were left.

3.1.5 The investigation was told that the radiology booking team had, on occasion, taken it upon themselves to use online translation services to relay appointment details to families via text. They had done this on an ad hoc basis, when they had telephoned a family and realised they did not speak English. This was not done in the case of the reference event, as the appointment reminder calls had not been answered and the radiology booking team had therefore not had this prompt to translate the appointment information.

3.2.1 The investigation found that the Patient in the reference event had been discussed in a weekly oncology MDT meeting, which had led to the MRI scan being arranged. These meetings were attended by oncology leads, consultants who had referred patients to the oncology meeting, a radiology consultant and support staff.

3.2.2 During the meeting, new patients with a confirmed oncology diagnosis were discussed, as well as those who had undergone an examination that week. The Patient in the reference event was included in the ‘for discussion’ section of the meeting, as they did not have a confirmed oncology diagnosis. As such, they were not included on the main MDT list of patients with a confirmed oncology diagnosis.

3.2.3 The investigation was told that an urgent (within two weeks) MRI scan was arranged to “rule out” cancer. Because it was thought unlikely that the scan would reveal cancer, the Patient was not transferred to the oncology team and instead remained under the care of the consultant who had taken the case to the MDT for discussion.

3.2.4 The oncology team told the investigation that there was no electronic system within the Trust to track patients who had been referred for an urgent MRI scan, but said they would not forget such patients due to being “concerned about them”. Additionally, the oncology team tracked patients with a confirmed oncology diagnosis using a spreadsheet, and explained that their usual practice was to “keep an eye” on the booking system to see if the scan had been arranged. If the booking system did not show an appointment within a week, the oncology team would contact the radiology booking team to find out what was happening. The oncology department also had a whiteboard that listed oncology patients who were awaiting scans.

3.2.5 In the reference event, the Patient did not have a confirmed oncology diagnosis and was not under the care of the oncology team. As such, they did not appear on the oncology team’s whiteboard or spreadsheet, and were not remembered by the team as someone for whom they were awaiting results.

3.2.6 The department of the consultant who had referred the patient for discussion at the MDT had no system in place to track patients for whom they were awaiting results. They had recognised this as a risk for patient’s details becoming lost but explained that there was no electronic system in place to support such tracking.

3.2.7 MDT coordinators are team members who, among other roles, ensure that actions arising from an MDT meeting remain on the agenda until they are completed. The oncology team in the reference event did not have a designated MDT coordinator. Instead, secretarial staff were undertaking elements of this role, with support from the clinical team. Staff explained that other oncology specialties had an MDT coordinator, and this department was an outlier in not having one. Staff said that the department had grown, but that an MDT coordinator had not been budgeted for or appointed. The lack of a designated MDT coordinator was considered by some staff to be instrumental in the reference event, as the Patient did not appear on the next MDT meeting agenda despite their MRI having not been completed.

3.3.1 Radiology booking staff said there were different booking procedures depending on the type of scan. Staff described having limited written operating procedures to follow, which did not outline all of the relevant steps. This meant that staff needed to use their experience, gained over time working in the team, to know the correct processes. Staff said that this situation had occurred as until recently the management structure had been cross site and the other site had needed more support, leaving processes to evolve without any oversight.

3.3.2 They said that this contrasted with some other departments within the Trust, such as the pharmacy, where processes were clearly written down. In radiology booking, “…it’s not…I just know.”

3.3.3 In the reference event, the MRI scan was requested as urgent and so followed the process for booking urgent scans.

3.3.4 Staff told the investigation that paper forms were commonly used for urgent scans, and that a discussion was needed between the requesting consultant and the radiology consultant before such scans could be booked. Scans were booked immediately following this discussion, after the radiology booking staff received the completed paper form.

3.3.5 Non-urgent scan requests were sent to the radiographers to decide on the number of time slots required for the scan and the imaging machine to be used. In contrast, urgent requests did not go to the radiographers. Instead, the booking staff were relied upon to know the correct number of time slots and the imaging machine that should be used, based on their knowledge.

3.3.6 Booking staff explained that this knowledge came with experience, and there was no system for checking their assessment. Staff felt this placed a lot of responsibility on the individual and that it “shouldn’t be like that”, with booking staff making an assessment that would be made by a radiographer for non-urgent scans. One booking staff member described that, when the patient had not previously had a general anaesthetic, it was their personal practice to speak with the pre-admission team and radiographers about bed booking and slot numbers, but that “everyone does everything differently” as there were no written procedures or guidelines on this.

3.3.7 When the radiology booking staff booked a scan for the child in the reference event, they contacted the medical secretaries to ask for a bed to be booked for a general anaesthetic, as the radiology booking staff did not have access to the bed booking system.

3.3.8 Once the scan and bed had been booked, the child’s family was sent a standard letter, in English, that included details of the appointment and the preparations required for the scan. The family was telephoned by the pre-admission and booking teams in relation to the appointment, as discussed in section 3.1.4. The calls were not answered and messages were left.

3.3.9 The investigation was told that there was no standard scan cancellation process. Rather, the process depended on whether the scan was cancelled by a member of the radiology booking team, the patient or another staff member. These processes were not supported by written procedures, were open to variation and had evolved over time. One booking staff member said that the processes were passed on by word of mouth, but that if you asked two staff members, they would likely tell you different things. Booking staff told the investigation that cancellations happened every day and could occur due to a variety of reasons, including:

• staff sickness
• non-urgent scans being postponed in favour of urgent ones
• the scan no longer being required
• the patient or their family cancelling due to illness or unavailability.

3.3.10 In the reference event, the child attended the appointment but had not been fasted prior to the scan, which was a requirement. This was discovered on their arrival at the hospital, and the ward nursing staff contacted the radiology booking team to explain why the cancellation was necessary and ask for the scan to be cancelled from the system.

3.3.11 The booking team was aware that the scan had been booked as urgent and told the investigation that, in such cases, they would contact the referring team to clarify if the scan remained urgent. Staff said that rebooking an urgent scan would require them to cancel another patient’s non-urgent scan and rebook them in turn. Staff explained that contacting the referring team was their usual practice at the time. This was borne out of experience and the lack of availability for rebooking scans, as opposed to being a documented procedure. The booking staff explained that after the incident of the child in the reference case, they changed their processes to routinely rebook urgent scans without contacting the referring team. They would ask the radiologists to decide which non-urgent scans could be cancelled to enable the rebooking of an urgent scan.

3.3.12 In the reference case, the booking team telephoned the secretarial team of the child’s consultant and asked whether the scan remained urgent. It was agreed that the query would be emailed to the secretarial team for action. An email was sent. No response was received in relation to the query for 11 weeks.

3.3.13 The investigation was told that booking staff relied on receiving a response from the secretarial team before rebooking appointments, but that there was no system in place to keep such queries highlighted. They explained that once an appointment had been cancelled, it was no longer possible to add notes to the electronic booking system to indicate that there was any follow up required by the team. Staff explained that this position contrasted with the previous electronic booking system, which had allowed entries after a booking was cancelled. They told the investigation that they had been given no input into the new system before it was put in place.

3.3.14 The team told the investigation that, since the reference event, individual members of the booking team had instituted their own methods of keeping track of areas where follow up was needed. For example, one staff member explained that if they were waiting for a response from a consultant’s secretary, they would print a copy of the referral form and keep it on their desk with a note attached, as a method of reminding them to follow up if necessary.

3.4.1 The radiology booking team was described by a senior member of staff as “the foundation of the department”. The team was designed to have four members but, due to people leaving, three members of staff were in post at the time of the reference event. On the day the initial scan was cancelled and not rebooked, only one member of the team was working.

3.4.2 At the time of the interviews for this investigation, the team comprised three full-time-equivalent members of staff. However, it was explained that because one person was new to the role, only two members of the team had knowledge of all of the booking processes. Of these two team members, one was due to leave the team for career progression, leaving only one person with overall knowledge of the systems. This staff member would be training any incoming staff on top of managing their own work. Training was described as “on the job”, so the quality of the training would depend on the experience of the staff, creating variations in practice.

3.4.3 The investigation escalated to the Trust its concerns about the staffing situation in the booking team. This was added to the Trust’s risk register and a full action plan was put in place to provide resilience to the service.

3.4.4 The radiology booking team was located in a designated office adjacent to the radiology secretaries. The secretaries were seated in an open office that provided access to the reception area and the consultants’ offices.

3.4.5 The investigation was told that the booking office environment was “stressful”, with staff constantly being “pestered”, distracted or interrupted.

3.4.6 The office was described as a “pain” because of the frequency of people walking in and out. When the office became busy with clinical staff, the booking staff felt unable to ask them to be quiet or leave because of the hierarchy, stating: “They are consultants. We are only band threes. You can’t talk like that to a consultant.”

3.4.7 Staff told the investigation that conversations with patients or families about cancelled scans could be challenging, as cancellations could be last minute or the patient might have been waiting for the scan for some time. Having these conversations in an office with interruptions and background noise could make them even more difficult for patients and staff.

3.4.8 The booking team described closing the office door and putting up a ‘meeting in progress’ notice to allow them to concentrate but said interruptions had continued. Staff also said they had asked for access to the corridor, where the booking office was located, to be restricted, but that this had been denied as it would have also restricted access to the consultants’ offices. The office disruptions had been escalated to the management team and a system of protected time was being trialled to try to tackle the issue.

3.4.9 There was significant cognitive load on the booking staff to manage a considerable workload while understaffed and using a system that relied heavily on individual memory.

3.4.10 While not used in the reference event, staff were keen to describe to the investigation the non-urgent booking process for an MRI, as this was a complex area in which patients’ details had previously been lost.

3.4.11 Non-urgent scan requests were made either on paper or via an electronic form. The electronic form had been created within the Trust and was completed on a system separate to the electronic booking system. Use of paper forms was discouraged by the radiology department, but staff said that some consultants insisted on using the paper version and the booking team did not feel able to enforce the use of electronic forms. Staff explained that this reluctance from consultants was in part due to more information being required on the electronic form than the paper one, meaning it took longer to complete. For example, the electronic form required the referrer to go through all the systems of the body and say whether they were relevant to the test being requested.

3.4.12 One clinician told the investigation that they were aware that electronic forms were preferred by the radiology department, and that these contained more information about the patient than the paper forms. They explained that the electronic forms took “10 times longer to complete”. They nonetheless said they preferred using an electronic form as it created an audit trail of the request within the electronic system, which could be referred to if necessary.

3.4.13 Paper forms were delivered to the booking staff by the referring team or the internal post system or were dropped at reception by patients. This meant that there were various ways in which the paper request form could reach (or fail to reach) the booking office staff.

3.4.14 If an electronic form was used, this would be completed by the referring clinician and emailed by the system to the radiology booking team’s inbox, as there was no interoperability between the system in which the referral was completed and the booking system. This meant that emailed referrals were printed and became part of a paper-based process.

3.4.15 In order for a non-urgent booking to be added to the system, the following events first had to happen:

• The booking approved by a radiologist.
• A radiographer confirmed the number of time slots required and what machine would be used.
• The pre-admission team undertook an assessment (for patients requiring a general anaesthetic).

3.4.16 For each of these steps to be completed, the same paper copy of the referral – together with a cover sheet for adding information at each stage – was moved around the department. This was the case whether the initial referral had been made on a paper or electronic form.

3.4.17 A spreadsheet was used to track the forms, as staff said they used to go missing during this process.

3.4.18 Staff said it could take 2–3 weeks before the process was complete and a scan could then be scheduled on the system. If a form was lost during the process – which would be identified by the spreadsheet – then the process had to start again as the handwritten information was not stored elsewhere.

3.4.19 In addition to the time taken to work through this process, there were numerous points where a patient’s details could be lost. The process relied on the experience and memory of staff, with no electronic system to support them.

4.1.1 All of the trusts visited told the investigation they provided interpretation services for face-to-face, telephone and video appointments. However, none of them routinely provided written communications, in relation to appointments, in languages other than English.

4.1.2 In 2012, as part of a 10-year strategy for transforming information for heath and care, the Department of Health stated:

‘For those of us who need support in accessing information, health and care professionals as “information givers” have a vital role. Professionals need excellent communication skills for sharing information and for communicating with us in ways that the person receiving care can understand. This includes thinking about language and interpretation support and ensuring that all communications are in formats that each of us – as the individual recipient of the care – can understand.’
(Department of Health, 2012)

4.1.3 The provision of non-English-language interpretation or translation (for reasons other than disability) is not included in the NHS England (2017) Accessible Information standard. However, one of the guiding principles of the NHS, as reflected in its constitution, is that: ‘The patient will be at the heart of everything the NHS does.’ As part of this, the constitution says that NHS services ‘must reflect, and should be coordinated around and tailored to, the needs and preferences of patients, their families and their carers’ (Department of Health and Social Care, 2021b).

4.1.4 Similarly, National Institute for Health and Care Excellence (2021) guidance on communications for patients with cancer states that appropriate information should be provided ‘in terms of language, ability and culture’.

4.1.5 The investigation identified that in addition to trusts not routinely translating written appointment details into a required language, there was no consistency in how booking office staff were made aware of the language needs of specific patients.

4.1.6 Staff from all levels within the trusts visited explained there was an expectation that patients would have a “family member or neighbour” translate letters for them. National guidance in relation to interpreting and translation services in primary care states that:

‘Reliance on family, friends or unqualified interpreters is strongly discouraged and would not be considered good practice.’
(NHS England, 2018)

While this is not replicated in the section on written language translation in this guidance, stakeholders agreed that the same principles applied.

4.1.7 Officers of the Equality and Human Rights Commission said that care providers expecting family members or neighbours to translate personal appointment and preparation letters may be considered to breach Article 8 of the Human Rights Act 1998 - people’s right to respect for their private life, their family life, their home and their correspondence (letters, telephone calls and emails, for example).

4.1.8 Some trusts told the investigation that they send patient appointment letters and information leaflets in English. The trusts said that patients had raised this as an issue. In response, steps were being taken to provide translated versions of information leaflets, although appointment letters had not been considered.

4.1.9 The investigation did identify some trusts that were taking steps towards addressing the issue of booking systems producing letters in English only. One trust said that it sends out translated appointment information to patients where needed, with the translations provided by a third-party supplier. Another trust described a policy whereby, for patients who they knew could not read English, no letter was sent but instead a telephone call would be arranged with an interpreter to talk through the details of the booking.

4.1.10 An additional trust had started to produce Braille appointment letters. This resulted from the action of a task force charged with improving equality and inclusion, and which aimed to include translations of letters in the future.

4.1.11 In 2022, Healthwatch noted that:

‘Getting to a hospital appointment is challenging when people don’t understand the appointment letter.’
(Healthwatch, 2022)

4.1.12 All of the appointment letters seen by the investigation were automatically produced by electronic booking systems. They were created using templates (selected from a list provided by the electronic booking system), which were populated with specific information about the clinical investigation. Some of the systems seen did not enable staff to view the letter before it was printed, although others did allow the letter to be reviewed in a word processor before printing. There were no options for appointment letters to be printed in other languages.

4.1.13 Staff explained that to produce letters in other languages, the system would need to have an option to select the relevant letter in a different language. As they understood it, this was a possibility as new template letters had been added to the system in the past.

4.1.14 The investigation identified translation tools that are used in other industries that need to communicate with customers or users. These include embedded translation functions within commonly used word-processing software, which translate a document as it is created. The investigation also found services such as gov.uk Notify, which provides communication output for public services, including the option for communications to be sent in different languages.

4.1.15 Additionally, the investigation noted the Written Medicine project, which has developed software to translate medicine labels into several languages. This was designed to improve patients’ understanding of their medications, increase compliance and reduce waste.

4.1.16 The investigation did not assess these tools in terms of the accuracy of their translations or their compatibility with systems currently used in NHS trusts. However, their existence demonstrates that it is feasible to translate written communications for patients. Translation enables more robust information sharing, while enhancing patient support and the patient experience.

4.1.17 There is a duty placed on doctors by General Medical Council (2013) and NHS England (2018) guidance, which similarly recognise the need for patients to receive information in a way they can understand. The NHS England guidance is directed at primary care providers, but the guidance notes that it may be applicable to other care settings, including ‘hospital sites’.

4.1.18 The Office for Health Improvement and Disparities Migrant Health Guide states:

‘If a patient requires interpreting and translation to access appointments and health-related information, they will also likely require interpreting and translation services for other communications with health services. This includes booking, cancelling and rescheduling appointments, and reading appointment letters’.
(Office for Health Improvement and Disparities, 2021)

4.1.19 The NHS Accessible Information Standard is intended to address ‘information or communication needs relating to a disability, impairment or sensory loss’ and specifically excludes non-English language interpretation and translation (NHS England, 2017). However, there is no comparable standard for non-English written communications. Some stakeholders considered that to not include community languages within this standard was a missed opportunity to ensure this area of inequality was highlighted and was something trusts were measured against.

4.1.20 The ‘Public Sector Equality Duty’ was created by the Equality Act 2010 and requires public bodies to have ‘due regard’ for advancing equality. This means:

‘[The equality duty] requires equality considerations to be reflected into the design of policies and the delivery of services, including internal policies, and for these issues to be kept under review.’
(Equality and Human Rights Commission, 2022)

4.1.21 NHS England told the investigation that there were future plans to scope the issues around community languages, including interpretation and translation. It was indicated that following the findings of the HSIB investigation this issue would be escalated to the NHS corporate risk register.

4.1.22 The investigation discussed its findings with Officers at the Equality and Human Rights Commission (EHRC). They said ‘NHS England and all care providers in Britain were covered by the public sector equality duty. As such, they should consider the impact their policies and decisions could have on people with different characteristics including race.’ They explained that ‘Race does not only cover colour, but it also includes nationality, ethnic or national origins’. The Officers considered that the HSIB investigation highlighted a ‘gap that needs to be remedied urgently’. They said that doing so would enable NHS England to demonstrate its compliance with the Public Sector Equality Duty, take actions to eliminate discrimination and to advance equality of opportunities. The EHRC also agreed to discuss this matter with the Care Quality Commission, with whom it shares a memorandum of understanding.

4.1.23 Some electronic patient record (EPR) systems seen by the investigation listed the patient’s nationality, but did not indicate their preferred language for communications. Other EPRs did have this option, but staff from different trusts reported that the information was not consistently entered. They explained that when a patient’s preferred language requirement was present on the EPR, this information had either been added locally or fed into the system by the NHS Personal Demographics Service (PDS). The PDS is designed to be the single source of patient information, such as their full name, date of birth, address and registered GP practice. The PDS has many fields that can be completed.

4.1.24 Some staff told the investigation that the information held by the PDS can only be updated by a GP, and that any information inputted locally would be overwritten when the system refreshed.

4.1.25 The investigation engaged with the lead product manager for the PDS at NHS Digital to explore this point. They explained that the PDS can be updated by any healthcare provider, but that when information is changed (such as the patient’s address), an information mismatch will be highlighted to the GP when they next open the record. The GP then needs to confirm the change for the PDS to be updated.

4.1.26 The lead product manager said that “time and trust” were issues for GPs in verifying the data: they have limited time to confirm details, and the source of the update might not be clear. This may lead to the GP dismissing the update, with outdated information therefore remaining on the PDS.

4.1.27 The investigation was told by a GP that there had been little training for GPs on the PDS and on the implications of “clicking ok” when asked to confirm updates to patient details. This had resulted in some GPs being reluctant to update information on the PDS as they were concerned about the impact on the patient’s records.

4.1.28 To mitigate this risk, NHS Digital is currently looking into ways for patients to update their own PDS information. Such updates would not require verification by the GP. However, to ensure that language requirements were correct, patients who have such needs would have to understand how and why they should update the system. Factors such as deprivation and digital exclusion may impact on some patient’s ability to engage with this.

4.1.29 The lead product manager explained that the PDS already contains a relevant field, labelled ‘preferred written communication method’. This can be used to record written language requirements and has been present on the system since 2005. The lead product manager said that barriers to completing this field include a lack of knowledge of its existence, as it does not form part of the administrative dataset (that is, basic details of name, address and so on) and is “difficult to find” in the system, and that there is no requirement for it to be completed.

4.1.30 The NHS England Accessible Information standard states:

‘Organisations MAY record details of individuals’ need for foreign language interpretation or translation alongside recording of information and communication support needs in line with the Standard, but this is optional.’
(NHS England, 2017)

4.1.31 Conversely, primary care guidance on interpreting and translation services states:

‘Language preferences and communication needs should be recorded in the patient’s record and shared with other services when the patient is referred on (for example to secondary care services).’
(NHS England, 2018)

4.1.32 More recent national guidance from the Office for Health Improvement and Disparities reinforces this point:

‘Record a patient’s language and interpreting needs in their healthcare record. Pass on this information when referring them to other health professionals.’
(Office for Health Improvement and Disparities, 2021)

4.1.33 The investigation found that when an entry is created in the PDS, the only mandatory field is the NHS number – which is created when the record is first opened. Along with this is an administrative dataset, which is also completed when the record is created. This information includes the patient’s name, sex, date of birth and so on. Ethnicity is included within this dataset, which may give some indication of language needs, but the investigation was told by NHS Digital that there is a requirement for ethnicity data inputted when a record is created to be removed from the PDS record after 6 months. The investigation was not able to establish the reason for this, and ethnicity data could be updated and retained by the system after that period.

4.1.34 In the trusts visited during the investigation, radiology booking systems pulled information from the PDS administrative dataset to populate patients’ details. Routinely completing the preferred written communication method field would mean that hospital systems that draw information from PDS would also display this requirement, where allowed by the hospital’s system.

4.1.35 The investigation found that staff could also become aware of a patient’s language requirements when:

• this detail was included in the clinical information from the referral
• telephone contact was made with the patient or their family.

4.1.36 Information on a patient’s individual needs is essential to ensure that communication occurs in a way that will be understood by that patient. Because the PDS integrates with many other healthcare systems, it is uniquely placed to ensure that healthcare providers and administrative staff are informed of a patient’s needs.

4.1.41 The investigation acknowledges that this report relates to safety and not to financial considerations. However, the investigation often heard concerns around the cost and feasibility of translating written communications, as well as the additional work for those working with the PDS in ensuring the language fields were utilised. Specifically, concerns were voiced over ensuring that clinical information has been translated correctly.

4.1.42 This investigation is limited to the letters produced by booking systems. The systems viewed by the investigation produced template letters with standard content. As such, a translation would need verifying only once, after which the translated letter could enter the system as an additional template.

4.1.43 The investigation engaged with a Trust that has made health equality and inclusion a key organisational priority. At the time of the discussion, the Trust had 14 formal workstreams on addressing disparities in healthcare. One such workstream, the ‘respiratory did not attend (DNA)’ pilot, was developed to increase patient attendance at respiratory appointments. A disparity had been identified between non-attendance rates for White British patients (4–8%) and ethnic minority patients (30–50%). A cohort of 492 ethnic minority patients were telephoned 2 weeks prior to their appointment, and offers were made to provide an interpreter or transportation to the appointment. Some of those contacted did not know of the planned appointment as they could not read the letter or text they had received. Following these interventions, the non-attendance rate for those contacted for the pilot was reduced to 8%.

4.1.44 The investigation was told that the saving in terms of clinical time far exceeded the investment required to reduce the non-attendance rate. In addition, the trust recorded each patient’s preferences for communicating information on their EPR, making attendance at future appointments more likely.

4.1.45 Encouraging attendance at appointments not only has an impact on clinic time, but would also be expected to have long-term benefits for patients through earlier detection and treatment of disease.

4.1.46 The investigation explored the impact of non-attendance with other trusts, and all said that non-attendance consumed resources and led to unused appointments in an already pressured environment. Trusts could not supply data on the number of appointments missed because of language barriers.

4.1.47 The investigation identified that the national healthcare inequalities landscape is confusing. This finding was echoed by NHS staff, including those who work within this area.

4.1.48 The investigation identified various teams within NHS England that have specific remits in relation to inequality. The investigation contacted the patient equality team at NHS England, who explained that the issue of translated appointment information was not within their remit as it did not relate to the Accessible Information Standard. Similarly, the National Healthcare Inequalities Improvement Programme at NHS England said that this issue did not fall within its current workstreams.

4.1.49 The investigation learned that the public sector equality duty (Patient Equalities Team), and health inequalities duties (Healthcare Inequalities Improvement Programme) teams are distinct, though related, and fall under the sponsorship of different National Executive teams within NHS England. However, the teams with day-to-day responsibilities for work on equality and health inequalities work closely together.

4.1.50 NHS England staff described the inequalities landscape as “fractured”, with no consideration of the patient journey as a whole or of those parts of the NHS with which patients who do not communicate in English will interact, including booking systems. The investigation heard that there was “a lack of ownership for issues of interpretation and translation [that goes] beyond the NHS into other public sector organisations”.

4.2.1 The investigation found that, in all of the trusts visited, radiology booking systems operated in a pressured environment. Pressure was exerted by:

• appointment requirements and capacity
• staff pressures.

4.2.2 At the end of July 2022, 6.8 million patients were waiting to start treatment, many of whom required diagnostic tests (The Royal College of Radiologists, 2022).

4.2.3 Some trusts told the investigation that they had a backlog of appointments and gave several reasons for this. One trust said that as the understanding of medical conditions increases, more types of scans are being undertaken using various machines with different capabilities and patients are being monitored more frequently. It has been recognised nationally that:

‘Radiology is being used earlier and more extensively in the diagnostic pathway. It is at the heart of a growing number of screening programmes and health checks.’
(Getting it Right First Time, 2020)

4.2.4 Staff considered that these issues were particularly relevant to radiology as it is “not in charge of its own destiny” because all of its work comes from referring clinicians, who are requesting scans more frequently.

4.2.5 Patients not attending for their appointments was also recognised as a problem that impacted upon capacity, as these appointment slots would go unused and the patient would have to be rebooked, adding to the workload of the department.

4.2.6 In contrast to the Trust in the reference event, the investigation found that some systems had dedicated times for ‘urgent’ and ‘routine’ scans, so as to ensure slots were available for urgent scans. The investigation witnessed a variety of methods for this, including colour-coding slots, which the booking staff felt worked well.

4.2.7 The investigation was told by a trust that it had the same number of scanners as it did 10 years ago, but that demand for scans had increased by 10% year on year. This meant that scans were being arranged for evenings and weekends, in addition to during core working hours. This situation is corroborated by the ‘Clinical radiology: UK workforce census 2020 report’ (The Royal College of Radiologists, 2021), which found that radiology clinical directors were concerned about insufficient scanners to meet demand.

4.2.8 None of the booking teams spoken to during the investigation were fully staffed. Some teams were running at 50% staffing, with this having a knock-on effect on staff wellbeing. Some staff said that they experienced significant stress in their job, and that some people had joined the team and left after a short time due to the pressures of the role. One booking team member said that they were sometimes on their own with responsibility for coordinating more than 100 referrals, and felt this was more than should be expected of a person in their role.

4.2.9 Staff reported distractions and interruptions affecting their ability to do their work to the best of their ability. In the reference event, the booking office staff said that conversations between consultants would often take place in the booking office and distract the staff. It has been recognised that:

‘Chronic noise exposure, even at moderate intensity, can hamper psychomotor, intellectual, attention, and memory functions. This impairment is especially apparent when an individual is exposed to “irrelevant speech” while engaged in mental activities that rely heavily on working memory, such as during multitasking.’
(Katz, 2014)

4.2.10 Additionally, some booking staff told the investigation that although their team was understaffed, they could still be redeployed to another area – such as reception – to cover absences. This compounded staffing issues within booking teams.

4.2.11 Some staff said that the pressures of the job and the banding of the role significantly affected recruitment. Administrative booking office staff are commonly Bands 2–4, with annual pay of £20,270–£26,282 (NHS Employers, 2022). Some trusts said they were exploring home-working options for administrative booking teams, as this allowed for a wider pool of applicants and reduced costs for staff by not having to commute. The investigation notes that home-working would also reduce the distractions referred to above (section 4.2.9).

4.2.12 In addition to a lack of administrative staff, many trusts reported not having a full complement of either radiologists or radiographers. This is a national problem, with the 2020 Royal College of Radiologists census (The Royal College of Radiologists, 2021) revealing an estimated shortage of nearly 2,000 consultant radiologists. This same report found that stress, burnout and low morale are widespread in this discipline.

4.2.13 The investigation found that staff responsibilities varied widely from one trust to another. At the reference event Trust, decisions on the scanner to be used were made by radiographers or administrative booking staff. In contrast, at other trusts, radiologists were the only decision-makers in relation to MRIs. Administrative staff were commonly relied upon to use their ‘experience’ to allocate booking slots, although one trust had clearly defined slots in the booking system for different types of test, negating the need for administrative staff to make such decisions.

Hybrid systems

4.2.14 The investigation carried out observations at multiple trusts and saw varying processes for radiology bookings. Some of the systems seen by the investigation gave the impression of full electronic integration. On closer inspection, the investigation found that paper elements were used to bridge gaps, thus creating hybrid systems.

4.2.15 An example of this, seen by the investigation, was an electronic system that – for scans booked within the Trust – allowed patients to be booked and tracked, and delivered the results to the referring clinician. However, for requests originating outside of the Trust, letters and forms were printed and placed in trays so that they could be reviewed before being manually added to the electronic system. This system can be seen in figure 1.

4.2.16 The tracking of these referrals relied upon paper being moved into the appropriate tray depending on whether they were requested, awaiting results or had issues that needed reconciling before they could be completed.

4.2.17 These gaps within the electronic system meant that it did not meet the needs of the user, and workarounds were put in place by staff to keep the service running as required. Additionally, the paper elements of the system relied on people inputting and tracking the papers, increasing the risk of variability.

4.2.18 The system described above used 10 reams (of 500 sheets) of paper each week. Staff indicated that they would prefer a single, fully integrated electronic system that did not require the use of paper.

4.2.19 The investigation was told that it was common for referring clinicians to not receive any feedback on the status of a scan request before receiving the scan report or, in some situations, notification that the patient did not attend.

4.2.20 Staff described some systems used to request radiology investigations as being overly complex and receiving little training on their use. Others considered that “digital illiteracy” (the inability to use digital tools) and “incomplete use of systems” among clinical staff affected their effectiveness. One example given was that when completing an electronic request for imaging, the clinician needed to select a particular patient interaction in the electronic patient record, such as a clinic appointment to make the request. This was known as an ‘encounter’. Once the imaging had been completed, the system would return the scan results to the clinician associated with the encounter in the request. A common issue with this system was that clinicians would select an encounter at random to allow them to make the request. As a result, the clinician making the request would not receive the result of the scan.

4.2.21 While this example was given to the investigation as one in which staff were incorrectly using a system, it demonstrates that the electronic system allowed variability and did not necessarily support the clinician.

User-centred design

4.2.22 All of the booking system processes described above have gaps that can result in patients being lost to follow-up. The gaps have occurred because the design of the systems has not properly considered all of the users and the routes through that system.

4.2.23 One booking office team told the investigation that previous systems at their Trust had more options available and were overall “better” than those that had been more recently introduced. Additionally, some systems had been purchased on the basis that they could integrate with existing systems – but, once in place, it had become clear that only limited integration was possible and that further adjustments to the system would result in development costs.

4.2.24 Staff said that they had no input into the change of systems. There had been no testing of the systems with them as end users, before the systems being implemented.

4.2.25 The Central Digital and Data Office, which leads the digital, data and technology function of the UK government, has produced guidance on the government’s design principles (Central Digital and Data Office, 2012). The first element of this guidance relates to user needs and states:

‘Service design starts with identifying user needs. If you don’t know what the user needs are, you won’t build the right thing. Do research, analyse data, talk to users.’
(Central Digital and Data Office, 2012)

4.2.26 The importance of user-centred design has been recognised by NHS Digital, which has expanded its user research team. The role of this team is to deliver and lead user centred design of NHS Digital products.

4.2.27 The investigation spoke with the Head of User Research at NHS Digital. They said that user centred design of staff facing digital products was evolving and historically not always used.

4.2.28 NHS England highlighted to the investigation Clinical Safety Standard DCB0160 which details the need to ensure any risks to individual patient groups are identified and appropriately mitigated. It states:

‘The Health Organisation MUST identify and document known and foreseeable hazards to patients in both normal and fault conditions through the introduction and use of the Health IT System’
(NHS Digital, 2018).

4.2.29 Additionally, the NHS Service Standard (NHS, 2022) was created to guide NHS Digital teams in complying with the gov.uk Service Standard on how to ‘create and run great public services’ (HM Government, n.d.). An assessment that the Service Standard is being met is currently only required for national NHS teams that are dealing with large volumes of transactions. For all NHS teams, the NHS Service Standard is deemed to be a ‘best practice’ guideline.

4.2.30 NHS Digital told the investigation that the NHS Service Standard is not always applied locally to external suppliers of electronic systems to NHS trusts.

4.2.31 The investigation spoke with a member of staff from the Department of Health and Social Care who said that while adherence to the Service Standard was not mandatory for frontline services, there was an intention for the standard to be applied more widely.

4.2.32 They also said that as part of the Service Standard all users of a system should be considered, and that “would include patients”. Therefore, in the example of a patient requiring written information in another language, such a ‘user persona’ (a fictional representation of a potential user) would need to be considered as part of the Service Standard.

• Written communications to patients about radiology appointments are routinely sent in English only.
• Healthcare staff expect that written appointment information will be translated by a patient’s friend or family member.
• NHS England standards do not require written appointment information to be given in any non-English language, other than for people with a disability.
• Confusion about the requirements for appointments can result in delayed care and additional costs if appointments need to be rebooked.
• The language needs of patients are not always clearly understood. There is differing national guidance as to whether written communication needs should be recorded.
• The national NHS system that holds patient information and populates it into many trust systems (the Personal Demographics Service) can store information on patients’ preferred written communication methods. However, this information is often not entered into the system.
• Administrative staff are not routinely involved in assessing and testing electronic booking systems before they are implemented.

The Chief Investigator authorised a national investigation based on HSIB’s patient safety risk criteria, as described below.

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

Patients being lost to follow-up can lead to missed or delayed clinical investigations, which may cause patient harm by delaying treatment.

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

HSIB used the Strategic Executive Information System (StEIS) to search for incidents in the ‘treatment delay’ category that included the keyword phrase ‘lost to follow’ (see table 1). Thirty-four incidents were identified. The delays in these 34 incidents appear to have been caused by issues with keeping track of patients, rather than by a lack of capacity to undertake clinical investigations. In several cases, the patients’ treatment options and prognosis were thought to have been adversely affected by the delay.

These incidents were reported across the country and across different disciplines, indicating that this is a widespread issue and not related to a single 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?

A national safety investigation can provide insight into persistent safety risks and make recommendations that stimulate change. In addition, HSIB investigations provide an opportunity to share learnings from stakeholders and/or healthcare providers who have made improvements to processes and practices. 

The investigation was completed between July 2022 and January 2023.

The investigation interviewed staff involved in the reference event and those from the wider organisation. Semi-structured interviews were undertaken face to face and virtually with:

• administrative booking office staff
• clinical staff involved in referrals for and feedback from radiological examinations
• leads for governance and digital systems.

The investigation visited the Trust involved in the reference event, and observed the systems and processes used in arranging, booking, cancelling and rebooking radiology appointments.

Two acute hospitals were visited during the national investigation, and a further acute hospital was engaged with remotely. The investigation met with administrative booking office staff, clinical staff and governance and digital systems leads for the hospitals. It also engaged with national NHS bodies and advisors in the areas being explored (see below). Further evidence was gathered from national policy and guidance, and the research literature.

The investigation used the Functional Resonance Analysis Method (FRAM) (Hollnagel, 2012) to explore the booking system processes for the reference event investigation. This provided insight into gaps and variations in the processes.

The investigation used the Systems Engineering Initiative for Patient Safety (SEIPS) (Carayon et al. 2006) to explore the wider national picture. This tool was used as a guide during site visits for collecting evidence and in analysing the data gathered. SEIPS provides a human factors framework for understanding structures, processes and outcomes, and the relationships between these.

The investigation engaged with stakeholders (see table 2) to gather evidence during the course of the investigation. This also enabled checking for factual accuracy and overall sense-checking. The stakeholders contributed to the development of the safety recommendation and safety observations based on the evidence gathered.