Recommendations but no action: improving the effectiveness of quality and safety recommendations in healthcare

Quality and safety recommendations are made to the healthcare system as a mechanism to drive improvements and/or mitigate an identified patient safety risk. These recommendations are made by many different stakeholders both within the healthcare system and outside of it, and can be directed towards any level of the healthcare system, for example national organisations or individual providers. This report distinguishes between recommendations and regulatory actions or requirements, the latter being outside of the scope of this work.

The intention of all such recommendations is to improve outcomes for people who use healthcare services and for staff working within healthcare. However, the sheer number being made and the variance in their quality means that they can be a burden to an already pressured healthcare system which is expected to digest, prioritise, pay for and implement actions in relation to them. This can lead to a lack of action in response to recommendations which means the improvement does not happen or the patient safety risk can remain.

This report is an output of the work commissioned by the Department of Health and Social Care (DHSC). At a meeting of the Arm’s-Length Body (ALB) Chief Executives and Chairs in November 2022, areas where joint work between DHSC’s ALBs would be valuable were identified. A series of workstreams were set up as a result.

Dr Rosie Benneyworth, the Chief Executive Officer of HSSIB, agreed to Chair the workstream on how ALBs and DHSC can better manage risks that the system is facing. This became known as the Recommendations to Impact Collaborative Group (referred to here as ‘the group’) and has been meeting virtually and at in-person workshops since March 2023. The group is a collection of organisations and individuals, including a panel of international academics and experts in collaborative governance and the role of evidence in developing policy. The purpose of these meetings is to look at ways in which to increase collaboration and efficiencies in how safety recommendations made to the healthcare system are developed, made and implemented.

This report is now being published by HSSIB on behalf of all ALBs who are part of this group.

The following organisations and individuals have contributed to this work:

• Failure to implement actions following recommendations can impact public confidence in the healthcare system and compound harm to patients.
• The ‘noise’ created by the significant volume of recommendations being made to the healthcare system means that providers struggle to prioritise and implement recommendations, concentrating on those which are addressed directly to the provider, or where there are immediate patient safety risks.
• Some recommendations duplicate or contradict others. The development of a searchable repository which includes recommendations made across the healthcare system may help to reduce this.
• It may reduce the ‘noise’ and help with prioritisation if organisations refer to each other’s recommendations, or group together in support of one organisation’s recommendation rather than repeating it. The development of an agreed system to identify recommendations for cross-referencing would assist this.
• There is currently a lack of visibility of ongoing work across arm’s length bodies that would enable collaborative working on related workstreams. A searchable repository of ongoing work may assist this.
• Recommendations differ in terms of the evidence on which they are based, and their structure and language. This can affect their relevance and how they are interpreted.
• It is unclear how some recommendations are intended to impact the patient, which should be a key consideration in their development where possible.
• Most recommendations made to the healthcare system are not costed, either in relation to the cost of implementing the proposed actions or their longer-term cost effectiveness. This may affect providers’ ability to implement them and means there is a lack of information to support prioritisation decisions.
• Some recommendations may be of limited relevance to certain providers and could promote inequalities by negatively impacting certain patient groups if implemented. However, providers can feel they are not empowered to reject recommendations, especially those related to safety.
• Few recommendations require a formal response from the recipient organisation, and there is a lack of monitoring of the actions planned or taken to address recommendations. A monitoring system could help to track actions and identify opportunities for escalation where changes have not been made.

The Recommendations to Impact Collaborative Group recommend further work in this area to develop:

Healthcare providers told the group that they were being “swamped” by the number of recommendations being made by many different organisations, and that implementation of all such recommendations was not possible or in some cases desirable. These recommendations were in addition to regulatory actions, which must be completed. Some providers described having received hundreds of recommendations which had been made by various organisations and said that it was difficult to know where to start with implementation. They also highlighted that some recommendations duplicated or conflicted with others, making it difficult for providers to know which ones to implement. The ‘burden’ of recommendations on providers was recognised by the investigation into maternity and neonatal services in East Kent, as was the risk that making additional recommendations ‘would inevitably repeat those made previously, or conflict with them, or both’ (Kirkup, 2022).

The group discussed the need for organisations to take a more pragmatic approach so that they avoid repetition, by referring to a similar recommendation made by another organisation, or grouping together to support one organisation’s recommendation rather than repeating it. The development of an agreed system to identify recommendations was also discussed, to create a simpler and more joined-up process for cross-referencing recommendations.

Sometimes recommendations are made to healthcare organisations that contradict with previous recommendations. The group identified that silo working (that is, organisations and/or the teams delivering workstreams working in isolation) was one cause of this. For example, it is not uncommon for different arm’s length bodies (ALBs) to be undertaking investigations or reviews within related areas of healthcare without knowledge of each other’s work. The group considered that a searchable repository (central online store) of ongoing work within ALBs would enable collaborative working on related workstreams, which may reduce duplication of recommendations and avoid contradictory recommendations. The group plans to develop a proposal for the creation of such a repository while keeping in mind the need to avoid additional administrative burden.

The group recognised that ALBs were just one set of organisations feeding recommendations into the healthcare system. Many recommendations are made by organisations with no statutory responsibility to do so. In some other industries, recommendations made by those without statutory authority may not be implemented. However, in the healthcare sector organisations are reluctant to not take action on recommendations because of the repercussions for patients and a desire to improve patient safety wherever possible.

The ‘noise’ created by a significant volume of recommendations has been recognised by other industries and has led to organisations such as the Air Accidents Investigation Branch (AAIB) changing their approach. The AAIB told the group that during the 1980s and 1990s it would produce hundreds of recommendations, many of which were quite direct in their language and the actions they expected organisations to take. It has since changed its philosophy on recommendations and now highlights the risk and outcome without specifying the solution, as the recipient is often best placed to develop this. Also, recommendations made previously which have been rejected are not repeated unless there is additional evidence to support them; in such cases the recommendation is reworded to reflect this.

Making reference to previously made recommendations is possible in the aviation sector, as there are repositories of recommendations that have been made across the sector which hold most of the recommendations for future reference.

NHS England has developed a National Recommendations Register for Maternity and Neonatal Services, which is available from its FutureNHS platform. This repository includes 708 recommendations from ‘published national maternity and neonatal reports and audits’ and allows users to search these by care setting. The development of this resource over a number of years by NHS England is recognition of the need for visibility of recommendations made. Similarly, the Health Quality Improvement Partnership, which commissions and contract manages the National Clinical Audit and Patient Outcome Programme (NCAPOP) has developed a publicly available recommendations repository. This includes recommendations made in reports published by the NCAPOP. It was developed following feedback from trusts on the recommendations ‘burden’ they were experiencing. Most recently, the Grenfell Tower Inquiry: Phase 2 Report (2024) recommends for the government to maintain a publicly accessible record of recommendations made by select committees, coroners and public inquiries together with a description of the steps taken in response.

However, there is currently no repository which includes recommendations made across the wider healthcare system, which the group considered would be useful. The group is therefore developing a proposal for such a repository.

The group found that because recommendations were being made by so many different organisations, the robustness of the evidence on which they were based, and the structure and language used, differed. This meant that individual providers had to interpret the recommendations, which they often did differently, potentially creating inconsistency. The intended meaning of the recommendation could also be lost if the wording was unclear.

Some recommendations made were also unclear as to the intended impact on the patient or whether there was consideration of the patient when the recommendation was drafted. The group agreed that a “line of sight” to the patient should be a key consideration in the development of recommendations.

In recognition of the inconsistency of recommendations the NHS England Patient Safety Team has developed a framework for the creation of recommendations. This structure uses the acronym CREATED SMART (see appendix b), a variation on the commonly used SMART approach, to highlight the components of and considerations for a recommendation. The idea is that this is to be used across NHS England for the development of recommendations. While uptake at this time is variable, it has been built into the framework for how the NHS England Patient Safety Team commissions independent investigations. Other organisations in the group used SMART or aspects of this framework to develop recommendations, while others used collaborative working groups or human factors principles (which focus on the interaction between people and the elements of the system in which they work). Some of the principles used overlapped but there was no consistent approach across the organisations in the group to the development of recommendations.

A consistent theme across the group’s discussions was the importance of taking a collaborative approach to the creation of recommendations. Some reports and national inquiries deliver recommendations at the point of publication without prior engagement with the intended recipients. The group learned that approaches to collaboration differed across ALBs but there was strong agreement that involving the intended recipient in the development of recommendations was the right approach.

The advantages of this were cited as including:

• assurance that the correct recipient had been identified with the remit to effect change
• it enables the opportunity to learn from mitigations which may have already been explored by the recipient
• it can improve the response to recommendations as they have been co-created
• it promotes a sense of ‘with, not to’, and
• it helps to ensure that the recommendation is realistic and within “the art of the possible”.

The importance of co-development of recommendations was said to be the same for those made at different levels of the healthcare system. For example, one trust told the group that recommendations following internal investigations were sometimes made by the investigator in isolation. This meant that the recommendation may not be practicable or realistic for the recipient team or department.

It was also highlighted that some recommendations that are made across the healthcare system may not be relevant for every provider. The group heard that some recommendations may have a negative impact on certain patient groups and drive inequality by shifting the focus of efforts and funds to an area which is not relevant to the patient population served by that provider. The group discussed this challenge and whether the providers felt, or should be, empowered to be able to reject recommendations which were not relevant to their patient population. Barriers to rejecting such recommendations included:

• it could be viewed negatively by staff and/or the public if an organisation did not implement something that was badged as an improvement to services or patient safety
• the provider being seen as defensive, or not being open and transparent.

Again, collaborating on and co-designing recommendations, and/or testing them, may be opportunities to identify unintended consequences.

A consistent structure and approach for use by accredited organisations exists in relation to National Patient Safety Alerts (NHS England, n.d.). However, these alerts are not recommendations but required actions. The development of an agreed approach and structure for recommendations, which can be used across different organisations, is something the group is currently working on.

Recommendations can only be effective if actions are implemented, and sustained. The group learned that due to the challenges associated with the volume and development of recommendations, it is not possible for providers to implement all the recommendations they receive. Instead they have to consider how to prioritise actions in response to recommendations.

Putting the onus on providers to prioritise the implementation of one recommendation over another has an impact on patients, carers and families, not only in relation to the safety and quality of care in future but also for those who have experienced a patient safety incident and want to see change through recommendations. This can compound the harm experienced and damage public confidence in the healthcare system. Additionally, the group heard about moral injury (psychological harm) caused to NHS staff where recommendations were not implemented and patient harm continued to occur.

In recognition of the challenge of prioritising recommendations, NHS Resolution has developed and piloted a tool to prioritise recommendations in emergency medicine, the support of which has been a focus of the group. Details of this work are included below, with additional information available from the Royal College of Physicians 'Commentary' magazine.

Some bodies which feed into the healthcare system require the recipient of recommendations to provide a response. These include national inquiries, HSSIB and coroners (in the form of Prevention of Future Deaths reports). Other organisations do not require or expect a response to recommendations made and most do not monitor or follow up on whether actions have been implemented as a result of the recommendation.

Indeed, even for recommendations which do require a response, such as those listed above, there is usually no process for monitoring or tracking what happens after the response is made, such as the actions taken. The lack of monitoring of such actions means it is not clear whether implementation has taken place, or, importantly, whether the actions have resulted in the required improvement to patient care. Such monitoring would also inform our understanding of what types of actions lead to positive change for patients.

The group heard that to patients, families or carers responses to recommendations can feel cold and defensive when for them the learning and change are aligned to the human impact which led to the recommendation. Additionally, the lack of monitoring of actions in response to recommendations can compound these feelings, affecting faith in the healthcare system.

The Parliamentary and Health Service Ombudsman undertook a survey of complainants to understand their perspective in relation to recommendations made following complaints about healthcare. Less than 40% of respondents prioritised a compensation payment as of high importance, while more than 80% felt that revising procedures to prevent the recurrence of the issue was of high importance. However, where there is inaction following a recommendation this can compound the harm caused. For example, the group engaged with a family who had experienced a coroner’s inquest following the death of their son, aged 22 months. At the conclusion of the inquest the coroner wrote a Prevention of Future Deaths report to six national bodies highlighting concerns. As is required, responses were received from these organisations. The family found these to be ‘generic’ and felt they were passing responsibility to another agency. Overall the family described being ‘extremely disappointed’ with the responses. Also, they had not realised that there was no mechanism to challenge a response to the coroner, should it be considered lacking or if it states that no changes will be made.

The House of Commons Justice Committee recognised this monitoring and implementation gap as part of its work on a follow-up inquiry into the Coroner Service. It stated that it would like to see ‘a cross-sector board within each sector overseen by the relevant Department of State, tasked with oversight and analysis of all relevant PFDs [Prevention of Future Deaths reports] and responses’ (Neill, 2024).

There is limited monitoring of implementation all the way through the healthcare system, including at government level, as evidenced by the recent reports of The Thirlwall Inquiry (2024) and the Expert Panel of the Health and Social Care Committee (2024).

Calls for the development of a structure for monitoring and oversight are not limited to healthcare. They have also been made in relation to deaths in state detention (Inquest, n.d). In addition, volume 1 of the recently published report of the Infected Blood Inquiry (2024) refers to the need for post inquiry scrutiny of the responses to, and potential implementation of, the recommendations made as a mechanism of accountability for the recipients of those recommendations. The creation of a monitoring structure and accountability mechanism for the implementation and evaluation of actions taken would be in line with established Safety Management System principles (Health Services Safety Investigations Body, 2023). Additionally, it would demonstrate a firm and public commitment to a learning culture following recurrent gaps in the implementation of recommendations designed to improve the safety of patients.

The group will explore and draft a proposal for a monitoring system with a multi-agency board which would feed into the Department of Health and Social Care to provide oversight and a route of escalation for recommendations when no actions have been implemented. In creating this proposal the group will consider the role which patients, families and carers could have on this board.

The group heard that the impact on patients of actions taken following a recommendation is not commonly known by the organisations that make recommendations, and few organisations have systems in place to try and identify such impact. All members of the group recognised the growing importance of being able to understand the impact of recommendations, to identify whether the actions taken in response to the recommendation have been effective. If they have not been effective there is then an opportunity to explore why and whether the recommendation needs to be redirected (to another organisation), reworded or reconsidered.

One challenge in relation to assessment of impact is that there may be some distance between the recommendation recipient and the actions being undertaken in healthcare services. There was recognition that balancing the reporting burden on healthcare providers with the need for assurance was difficult. Local ‘ownership’ of recommendations (by integrated care boards) was discussed as they are well placed to gather information about the implementation of actions, including those which span multiple providers and pathways, and the impact on patients within local providers. However, it was recognised that this was an aspiration rather than current practice. It was clear that ensuring accountability for the implementation of actions in response to national recommendations would require building the capacity to do so where required (locally within integrated care boards or nationally within organisations, or separately).

A first step towards the assessment of impact, discussed by the group, could be the inclusion of metrics within the wording of recommendations to indicate how impact is to be measured. This will be considered as part of the development of an agreed recommendations approach and structure, as discussed above.