Skip navigation | Accessibility and accesskey details | Sitemap

National Audit Office Value for Money Report: Executive Summary

A Safer Place for Patients: Learning to improve patient safety

Executive summary

  1. Every day over one million people are treated successfully by National Health Service (NHS) acute, ambulance and mental health trusts. However, healthcare relies on a range of complex interactions of people, skills, technologies and drugs, and sometimes things do go wrong. For most countries, patient safety is now the key issue in healthcare quality and risk management. The Department of Health (the Department) estimates that one in ten patients admitted to NHS hospitals will be unintentionally harmed, a rate similar to other developed countries. Around 50 per cent of these patient safety incidents [Footnote a] could have been avoided, if only lessons from previous incidents had been learned. Figure 1 details some of the key facts.






  2. There are numerous stakeholders with a role in keeping patients safe in the NHS, many of whom require trusts to report details of patient safety incidents and near misses to them (Figure 2). However, a number of previous National Audit Office reports have highlighted concerns that the NHS has limited information on the extent and impact of clinical and non-clinical incidents and trusts need to learn from these incidents and share good practice across the NHS more effectively (Appendix 1).


     
  3. In 2000, the Chief Medical Officer’s report An organisation with a memory1, identified that the key barriers to reducing the number of patient safety incidents were an organisational culture that inhibited reporting and the lack of a cohesive national system for identifying and sharing lessons learnt.
     
  4. In response, the Department published Building a safer NHS for patients3 detailing plans and a timetable for promoting patient safety. The goal was to encourage improvements in reporting and learning through the development of a new mandatory national reporting scheme for patient safety incidents and near misses. Central to the plan was establishing the National Patient Safety Agency to improve patient safety by reducing the risk of harm through error. The National Patient Safety Agency was expected to: collect and analyse information; assimilate other safety-related information from a variety of existing reporting systems; learn lessons and produce solutions.
     
  5. We therefore examined whether the NHS has been successful in improving the patient safety culture, encouraging reporting and learning from patient safety incidents. Key parts of our approach were a census of 267 NHS acute, ambulance and mental health trusts in Autumn 2004, followed by a re-survey in August 2005 and an omnibus survey of patients (Appendix 2). We also reviewed practices in other industries (Appendix 3) and international healthcare systems (Appendix 4), and the National Patient Safety Agency’s progress in developing its National Reporting and Learning System (Appendix 5) and other related activities (Appendix 6).

Overall conclusion

  1. An organisation with a memory1 was an important milestone in the NHS’s patient safety agenda and marked the drive to improve reporting and learning. At the local level the vast majority of trusts have developed a predominantly open and fair reporting culture but with pockets of blame and scope to improve their strategies for sharing good practice. Indeed in our re-survey we found that local performance had continued to improve with more trusts reporting having an open and fair reporting culture, more trusts with open reporting systems and improvements in perceptions of the levels of under-reporting. At the national level, progress on developing the national reporting system for learning has been slower than set out in the Department’s strategy of 20013 and there is a need to improve evaluation and sharing of lessons and solutions by all organisations with a stake in patient safety. There is also no clear system for monitoring that lessons are learned at the local level. Specifically:
    1. The safety culture within trusts is improving, driven largely by the Department’s clinical governance initiative4 and the development of more effective risk management systems in response to incentives under initiatives such as the NHS Litigation Authority’s Clinical Negligence Scheme for Trusts (Appendix 7). However, trusts are still predominantly reactive in their response to patient safety issues and parts of some organisations still operate a blame culture.
       
    2. All trusts have established effective reporting systems at the local level, although under-reporting remains a problem within some groups of staff, types of incidents and near misses. The National Patient Safety Agency did not develop and roll out the National Reporting and Learning System by December 2002 as originally envisaged. All trusts were linked to the system by 31 December 2004. By August 2005, at least 35 trusts still had not submitted any data to the National Reporting and Learning System.
       
    3. Most trusts pointed to specific improvements derived from lessons learnt from their local incident reporting systems, but these are still not widely promulgated, either within or between trusts. The National Patient Safety Agency has provided only limited feedback to trusts of evidence-based solutions or actions derived from the national reporting system. It published its first feedback report from the Patient Safety Observatory in July 20055.

The culture within NHS trusts is now more open and fair

  1. A just and fair culture is a key requirement if reporting and learning are to be improved. All trusts have continued to build on and develop their clinical governance arrangements, but with varying degrees of success. Most trusts have succeeded in reducing the blame culture. By helping trusts to deal more effectively with poorly performing doctors, the National Clinical Assessment Authority [Footnote b] is continuing to contribute to the development of a more open and fair culture and, as a result, suspensions have increasingly been avoided. However, the support provided applies only to doctors. In 2004, the National Patient Safety Agency produced guidance aimed at supporting trusts in assessing their safety culture and promulgated a tool to prompt trusts to focus on why the patient safety incident happened, and not who was to blame, and to adopt a systematic approach to decisions about the employee involved (Appendix 6).
     
  2. Within local organisations strong leadership and governance at chief executive and board level is crucial. Virtually all chief executives provided examples of their personal involvement with the patient safety agenda. Since 2001, over 130 trust boards or key members of trust boards have engaged with the Board Development Team at the NHS Clinical Governance Support Team. More recently non-executive trust board members from 113 trusts have undertaken Leadership in Patient Safety Training provided by the NHS Appointments Commission and the National Patient Safety Agency.
     
  3. An organisational top down approach on its own is not sufficient. The regulatory bodies, Royal Colleges and the other professional bodies have all placed greater emphasis on individual responsibility and accountability for patient safety. Although few trusts provided incentives for staff to improve patient safety, 93 per cent involved them in identifying priorities and designing solutions.
     
  4. As nine out of ten NHS employees work in teams6, effective communication between staff is important to reduce the risk of unintended harm to patients, yet trusts often cite failure in communication as a reason for an incident. Communicating openly with patients and carers is also essential but only 24 per cent of trusts were routinely informing patients when an incident that they had been involved in was reported to the trust.
     
  5. To provide evidence that NHS organisations were doing their reasonable best to manage themselves so as to protect patients, staff and the public against risks of all kinds, the Department established the mandatory Controls Assurance Standards in 1999. Trusts had to undertake a self-assessment against defined criteria. For the Risk Management System standard these criteria included board accountability, adverse incident reporting and complaints and claims handling. Over the five years of its operation average compliance increased from 52 per cent to 87 per cent.
     
  6. In August 2004, the Department announced that key elements of the Controls Assurance Standards would be incorporated in a new performance assessment framework based around a set of core and developmental standards (Standards for Better Health), with compliance evaluated by the Healthcare Commission. Safety is the first of seven domains in these standards (Appendix 7).
     
  7. Assessment of trusts’ risk management systems undertaken on behalf of the NHS Litigation Authority has also provided a strong incentive for trusts to improve their reporting and learning systems (Appendix 7). Each year since the operation of the Clinical Negligence Scheme for Trusts many trusts have gradually improved their risk management systems and seen their contributions reduced according to the level of compliance achieved (Figure 4 ).

Local reporting has improved but there have been delays in establishing an effective national system

  1. Unless trusts are confident that their reporting systems identify the main risks to patient safety they cannot target interventions effectively. All trusts had implemented integrated reporting systems as part of risk management. By 2005, the majority of these reporting systems were either confidential (34 per cent) or open (63 per cent) with 38 per cent of these trusts also providing an anonymous reporting route for use by staff who may be fearful of raising their concerns. Reported incidents were analysed at the local level with relevant information passed onto one or more of around 30 organisations.
     
  2. Seventy-eight per cent of trusts told us that their emphasis on encouraging reporting was having a positive impact on the number of incidents reported and the total number of patient safety incidents reported within trusts has risen year on year. Despite the general increase in reporting, trusts acknowledged that a substantial number of incidents still go unreported (trusts on average estimated that 22 per cent of incidents go unreported, mainly medication errors and incidents leading to serious harm). Reporting of near misses was also low, mainly due to different perceptions of what constitutes a near miss. Training can help improve levels of reporting but there has been no evaluation of the efficacy of courses and no system for accrediting those currently in use.
     
  3. Healthcare organisations in other countries, having compared the merits of anonymous and confidential reporting, have generally opted for confidential reporting. The Department proposed a confidential scheme, mandatory for trusts, to record patient safety incidents and near misses across the NHS3, however the National Patient Safety Agency recommended the development of two reporting systems, one which would interface with trusts’ incident reporting systems, but with the identity of the patient and person reporting stripped out, and the second, a totally anonymous voluntary e-Form which can be shared with the trust if the person making the report agrees.
     
  4. The roll out of the National Patient Safety Agency’s National Reporting and Learning System has taken two years longer than originally envisaged. By 31 December 2004 all trusts had the technology to link to the system but many still had to map details from their local system to the national system. By the end of March 2005, some 170 acute, ambulance and mental health trusts had reported 79,220 incidents (a further 6,122 incidents were reported by primary care trusts making a total of 85,342 patient safety incidents reported to the National Reporting and Learning System up to March 2005).
     
  5. The e-Form was launched in September 2004 and by April 2005, 108 reports had been made using this route. Whilst the National Patient Safety Agency does not know how many staff will make use of the e-Form, it believes this is a rich source of information for learning5 and provides a safety net for those who are too frightened to report to their local system. Five trusts told us that they do not want an anonymous system as this undermines local reporting and learning and that they would discourage use of the e-Form. Twenty-nine trusts are actively encouraging the use of the e-Form.
     
  6. Building a safer NHS for Patients3 required the NHS to establish agreed definitions of incidents for the purposes of reporting, gradually moving to an international standardised taxonomy (description and classification of incidents). The National Patient Safety Agency developed its taxonomy in consultation with trusts but it is unlike many trusts’ taxonomies and, in order to link to the national system, trusts had to map it to their own. At the time of our survey 82 per cent of trusts had had difficulties with the mapping exercise, and 17 per cent of these said that they had experienced major difficulties. Two-thirds of trusts told us that the national taxonomy was not specific enough so were continuing to use their own. It is also different from taxonomies used in other countries. The World Health Organisation is currently developing an international taxonomy which would require the National Patient Safety Agency and trusts to make changes to their taxonomies if they are to comply.
     
  7. The National Patient Safety Agency worked with the Medicines and Healthcare products Regulatory Agency in order to test the feasibility of a single data entry point for reports of errors involving medical devices. However, this did not prove possible due to the statutory responsibilities of the Medicines and Healthcare products Regulatory Agency and the requirements of the National Patient Safety Agency. Indeed there has been no further development in this area and trusts are still required to report the same incident to more than one organisation.
     
  8. Given that the Department’s aim was to encourage reporting, no targets were set for reducing the number of reported patient safety incidents. Rather, the Department set targets for reducing the incidence of four specific types of errors (maladministration of spinal injections; serious error in the use of medicines; suicides by mental health inpatients as a result of hanging from non-collapsible rails and harm in obstetrics and gynaecology). Whilst there have been no reports of incidents involving the first type of error, there are limited data to judge whether the target on medication errors has been realised and mixed messages on progress against the targets on suicide as a result of hanging and obstetrics and gynaecology. For example, although negligence claims for obstetrics and gynaecology appear to be reducing, the Healthcare Commission highlighted concerns about the safety of some maternity services7, 8.

A number of local and national systems are in place for analysing and sharing lessons learnt, but most are under-used

  1. Most trusts did analyse incident reports and other information. Indeed most had been carrying out in-depth investigations of incidents at the local level for a number of years. Seventy-six per cent of trusts told us that they were now encouraging staff to use the National Patient Safety Agency’s root cause analysis tool, with many noting that it had helped to improve the quality and consistency of in-depth investigations. A number of trusts remarked that monitoring and investigating incidents created additional demands on busy senior staff, and consequently they did not always conduct a full root cause analysis of all serious incidents. The quality of reports on investigations was also very variable and recommendations were rarely actioned by organisations outside the trust in which the event had occurred.
     
  2. Dissemination of learning and the development of solutions was patchy and there was also no systematic monitoring to ensure implementation within the trust. Clinical audit can be an effective way to evaluate whether improvements are being implemented but a number of National Audit Office reports have highlighted concerns about the limited extent and coverage of clinical audit (Appendix 1). The Commission for Health Improvement reported in 2004 that this was still under-developed in many trusts9.
     
  3. Over half of trusts reported that patients were involved in both identifying safety priorities and developing ways to prevent recurrence. However, only six per cent of patients we surveyed said they were consulted about how the safety incident they experienced could be prevented from happening to someone else.
     
  4. Ninety-nine per cent of trusts identified specific interventions that they had developed to address patient safety issues (some are described in this report). However, few trusts have carried out any cost benefit analysis of interventions/solutions to improve patient safety. Given the estimated £2 billion cost of extra bed days due to incidents and the potential litigation costs, we consider that in many circumstances the cost of intervention is likely to be far less than the cost of failing to prevent the incident.
     
  5. At a regional level, half the strategic health authorities used clinical governance networks to disseminate learning and in some areas they have introduced patient safety learning sets. However, a number told us they were ill-equipped to share lessons and many felt that they did not have the capacity or capability to monitor the implementation of good practice. There is also a risk that as foundation trusts are not required to report to strategic health authorities they will miss out on the sharing of learning. Other sources of learning are organised networks, like those for cancer and coronary heart disease, and ambulance trusts use the Ambulance Service Association. Since summer 2004, the National Patient Safety Agency’s 28 Patient Safety Managers have been working with most trusts to help share good practice.
     
  6. One way of disseminating information about necessary changes is the Department’s Safety Alert Broadcast System. The Department, the Medicines and Healthcare products Regulatory Agency, NHS Estates [Footnote c] and the National Patient Safety Agency issue safety alerts to trusts for them to act upon within a defined timescale. During 2004-05, trusts received 93 alerts through the System. Trusts told us that there was a lack of clarity in the rationale for the decision to release information as an alert and some felt that a number of these alerts did not tell them anything new. All wanted better links and communication between the bodies that issue notices via the Safety Alert Broadcast System. The Chief Medical Officer’s annual report10 identified concerns that compliance with alerts was slow and some trusts which reported compliance were subsequently found to be non-compliant.
     
  7. The Department expected that the new national reporting system for learning would bring about changes at trust and national levels, through the analysis of incidents and then subsequently their root causes. As at April 2005, the National Patient Safety Agency had issued limited feedback to trusts of lessons emerging from their reports to the national system. Although the National Reporting and Learning System has the capacity to collect contributory factors, these are not mandatory and the intention is to identify trends that can then be analysed in greater detail. Trusts told us they were concerned that information flow was one-way to the National Patient Safety Agency and the general perception was that the National Reporting and Learning System was simply an information collection system. The July 2005 report from the Patient Safety Observatory should start to address this perception.5
     
  8. The Department envisaged that the National Patient Safety Agency would assimilate other safety related information from a variety of existing reporting systems and other sources such as NHS complaints, litigation, National Confidential Enquiries and national audits (Figure 3). We found that there has been limited progress on assimilating and disseminating lessons from these different sources of information. Furthermore, the individual organisations responsible for litigation and complaints have until recently not made as much use of the valuable data they collect as they might to help trusts avoid similar incidents.


     
  9. The National Programme for Information Technology in the NHS, being delivered by the Department’s agency NHS Connecting for Health, has a crucial role in developing the technology to ensure that relevant information can be stored securely and accessed readily. A key component, the National Care Record, has significant potential to improve safety as lost or poorly completed records are a major contributory factor to patient safety incidents. Technology will also facilitate retrospective audits, improve access to guidance and reduce the risks of incorrect drug prescribing and dosages. In time, trusts’ individual reporting systems will be integrated into the National Programme. The National Patient Safety Agency is working with NHS Connecting for Health’s patient safety sub-group to take this forward.

CONCLUSIONS AND RECOMMENDATIONS

For the Department:

  1. The Department established a number of arm’s length bodies with a role in keeping patients safe. The Department needs to use its arm’s length bodies’ performance monitoring system to establish appropriate actions and milestones to:
  1. The National Clinical Assessment Authority has played a key role in improving the management of suspensions of doctors but other clinical staff are not covered by the Authority’s remit. In the Government’s response to the previous Committee of Public Accounts recommendation [Footnote d] to consider extending the Authority’s remit the Department told the Committee that the functions of the National Clinical Assessment Authority were being transferred to the National Patient Safety Agency from 1 April 2005, and that this consideration was therefore on hold. Given that the transfer is now complete, the Department should now respond fully to the Committee’s recommendation to consider extending the role of the National Clinical Assessment Service to other clinical staff.
     
  2. It is imperative that patient safety becomes a core part of professional training, including helping clinical staff understand their responsibility for patient safety and the benefits of working in an open and questioning environment. The Department needs to build on its work with the professional regulatory bodies and Royal Colleges to better embed patient safety training in all pre-registration professional training curricula and to raise the profile of patient safety issues in post-registration training.
     
  3. Despite the rationalisation envisaged in Building a safer NHS for patients3, trusts are still required to report the same incident to numerous national bodies and revise their data sets to capture new information which those bodies require. Wherever possible, incidents should only be reported once and, as trusts move to electronic reporting, the Department should explore the possibility of recommending a single entry point, for example via the National Programme for Information Technology in the NHS. As a minimum the Department should consult with NHS Connecting for Health, the NHS Health and Social Care Information Centre, the National Patient Safety Agency, the Medicines and Healthcare products Regulatory Agency and the relevant signatories of the Healthcare Inspection Concordat [Footnote e] to identify the scope to rationalise the number of data entry points.

For the National Patient Safety Agency:

  1. Many trusts and organisations involved in collecting data on patient safety incidents consider that the taxonomy developed by the National Patient Safety Agency is not specific enough for their purposes. The National Patient Safety Agency should work to adopt a taxonomy that ideally corresponds to the international taxonomy being developed by the World Health Organisation, but as a minimum should gain buy-in from all trusts and other bodies requiring reports on incidents to a mandatory minimum data set to ensure that there is consistency in the data collected at local and national levels.
     
  2. Many trusts are questioning the value of sending data to the National Reporting and Learning System given the lack of feedback and would like to see more of an emphasis on solutions. The National Patient Safety Agency needs to agree with the Department a regular publication timetable, so that opportunities to sensationalise the data are reduced, and provide examples of how the NHS is learning from the data. One option is to produce quarterly updates so that it becomes standard. The National Patient Safety Agency needs to expedite its evaluation and feedback programme and focus on developing solutions to nationwide problems to mitigate the risk that trusts will stop sending data to the National Reporting and Learning System. These solutions should be accompanied by a sample business case which trusts can then customise.
     
  3. There is little dissemination of learning between most trusts. The National Patient Safety Agency’s Patient Safety Managers should establish formal systems to capture learning in specialties and share learning across other teams and trusts at both local and national level. In addition they should investigate the possibility of establishing local networks similar to those for cancer, which will have the potential to improve the delivery of patient-centred care by disseminating learning about the whole patient journey.
     
  4. There is currently no scheme for accreditation or benchmarking of patient safety training; thus trusts have no assurance that the training they commission is a good product. The National Patient Safety Agency should look to other industries and together with the NHS Institute for Learning, Skills and Innovation, develop an accreditation scheme for all patient safety training supplied by external providers. It should also evaluate training programmes operated by trusts to build up a library of good practice to enable trusts to customise their training to best effect.
     
  5. NHS Connecting for Health has asked the National Patient Safety Agency to help assure the specification for the National Programme for Information Technology in the NHS to ensure that patient safety is inherent throughout the system. In taking this forward the National Patient Safety Agency should ensure that Connecting for Health fully understands and builds on the lessons from the development and roll out of the National Reporting and Learning System.

For the Healthcare Commission

  1. Safety alerts are an important mechanism for implementing solutions and we support the conclusions in the Chief Medical Officer’s recent report10. Information on compliance should be made public and the Healthcare Commission should place special focus on verification of NHS trusts’ compliance when assessing performance against the Standards for Better Health.
     
  2. No single NHS organisation is responsible for auditing implementation of best practice solutions for patient safety issues. The Healthcare Commission should ensure that in assessing the safety domain it builds in assessment criteria that evaluate how well solutions have been implemented.
     
  3. Information from complaints and litigation is still greatly under-exploited as a learning resource. The Healthcare Commission needs to expedite its in-depth analysis of information from the NHS Complaints system and share lessons on a regular basis. The Healthcare Commission needs to work with the NHS Litigation Authority and the National Patient Safety Agency to agree how best to share the data and where the responsibility lies for identifying key lessons and providing trusts with feedback from these analyses.

For NHS acute, mental health and ambulance trusts

  1. Despite improvements in safety culture many NHS employees still fear blame or unequal treatment if they report incidents and this remains a major barrier to increasing accurate and honest reporting. There is a need for trusts to re-enforce their commitment to an open and fair reporting culture and to support staffing initiatives to improve. Trusts should assess their safety culture using one of the established tools, such as those listed in the Seven steps to patient safety11, and implement an action plan to address the identified issues.
     
  2. Financial problems and staff shortages can push patient safety down the list of trusts’ priorities. Although the potential avoidable costs of patient safety incidents is estimated to be as much as £1 billion1, some areas of investment are likely to have a bigger pay back than others. Trusts should ensure that funding for managing and improving patient safety reflects the organisation’s risk register, and require their patient safety leads to develop annual business cases that demonstrate the opportunity costs of the improvements they plan to make, where relevant these should build on the solutions and accompanying business cases developed by the National Patient Safety Agency.
     
  3. Patients have little involvement in the identification of patient safety priorities and in the design of solutions in most trusts. Trusts need to engage patients more in identifying important patient safety issues and designing solutions and make better use of information gained through contacts with Patient Advice and Liaison Services. Trusts should ensure that they fully investigate complaints and litigation claims and analyse trends in both so as to learn from them.
     
  4. Under-reporting is a problem in some staff groups more than others and there is a perception amongst staff that not all employees take responsibility for patient safety reporting. Trusts should target specific training and feedback on those groups of staff that are less likely to report. They should liaise with the National Patient Safety Agency to identify and learn lessons from trusts which have achieved high reporting rates and also to build on the lessons from national initiatives to encourage reporting such as the work being done by the Agency on encouraging junior doctors to report.
     
  5. Near misses are generally under-reported and information on outcomes, particularly death and serious harm is poor. Trusts should ensure that their reporting policies clearly define a ‘near miss’ and should develop strategies to encourage more staff to report them to make sure potential serious incidents that were prevented are not overlooked. Trust should also triangulate information from various data sources such as complaints, claims, coroners reports etc to ensure that all deaths and serious harm as a result of a patient safety incident are recorded on their incident reporting system.
  1.  [back] Terminology developed by the National Patient Safety Agency to be used instead of the terms ‘adverse event’ or ‘clinical error’.
     
  2.  [back] The National Clinical Assessment Authority was established as a special health authority in 2001 to provide support and expert advice and an assessment service to NHS organisations that are faced with concerns over the performance of individual doctors and dentists (Appendix 2 details our previous work on this issue). Following the Department’s Arm's Length Bodies Review, from April 2005, the National Clinical Assessment Authority became part of the National Patient Safety Agency and was renamed the National Clinical Assessment Service.
     
  3.  [back] NHS Estates, responsibilities for health and safety environment alerts are being transferred to the Department as part of the Arm’s Length Bodies Review.
     
  4.  [back] The management of suspensions of clinical staff in NHS hospitals and ambulance trusts in England ; Forty-seventh Report of Session 2003-04 (HC 296).
     
  5.  [back] A Concordat between the Healthcare Commission and nine other bodies inspecting, regulating and auditing healthcare was launched in June 2004. The aim was to reduce overlap and duplication of inspection, improve co-ordination, support improvements in quality and make inspections proportionate transparent and accountable.