Article Text

This article has a correction. Please see:

Managing the after effects of serious patient safety incidents in the NHS: an online survey study
  1. Anna Pinto,
  2. Omar Faiz,
  3. Charles Vincent
  1. Division of Surgery, Department of Surgery & Cancer, St Mary's Campus, Imperial College London, London, UK
  1. Correspondence to Anna Pinto, Division of Surgery, Department of Surgery & Cancer, Imperial College London, St Mary's Campus, Norfolk Place, London W2 1PG, UK; a.pinto{at}


Objective To examine the current state of practice in English NHS Trusts in relation to the communication of serious patient safety incidents to patients and families and support for all parties involved.

Design Cross-sectional design using online surveys.

Participants 209 patient safety managers, responsible for the ‘Being Open’ policy (54% response rate).

Setting English NHS Trusts.

Results 98% of the participants reported that they are familiar with the National Patient Safety Agency Being Open guidance and 82% that they implement it more than half the time when incidents occur. However, provision of timely information was not reported as routine, with two-thirds of the discussions taking place 3–6 weeks after the investigation. The frequency of taking responsibility for harm was low for incidents of different severity levels but significantly lower for less serious ones (p<0.001). Follow-ups of patients in the long term and ex gratia payments were provided less than half the time. The most highly rated barriers to being open were clinical staff's fear of negative reactions from patients or their families and anxiety about litigation. Support practices for staff, such as debriefing and training on being open, were acknowledged as highly important but not always available.

Conclusions Awareness of the importance of being open is high among patient safety managers in English NHS Trusts, but there is still considerable scope for improvement in the management of the after effects of patient safety incidents. More research is needed on patients' and healthcare professionals' preferences for support after patient safety incidents.

  • Patient safety
  • adverse events
  • being open
  • second victim
  • risk management

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Patients and families may suffer in two ways after serious incidents, first from the incident itself and second from the way it is handled by the healthcare organisation concerned.1 ,2 It is now well established that when patients are harmed they and their families need an explanation, an apology, an assurance that action will be taken to prevent recurrence and, in many cases, additional treatment and support of various kinds.3–9 Concealment, evasion and lack of prompt action to ameliorate physical and psychological damage can cause immense distress and anger and lead to a breakdown of trust between families and staff.1 ,10 In June 2010, a public enquiry was announced in the UK into Mid Staffordshire NHS Foundation Trust, following reports by an independent regulator on severe failings in emergency care and patient concerns about the care they or a relative received at the Trust. One of the most striking findings from the review was the poor care given to some patients and the lack of openness after patients were harmed. The chair of the enquiry, Robert Francis, commented that ‘I heard of too many cases in which patients had experienced some traumatic event … after which the next of kin only discovered what had happened on arriving to visit. Common humanity requires the close family to be told as soon as possible of distressing events so that they can offer comfort and support’.11

Considerable progress has been made in the development of policies to encourage and support organisations in providing effective support for patients, families and also for staff, who may also be affected by serious incidents as existing literature suggests.12–18 A number of government agencies and professional bodies in the UK and internationally have published guidelines and statements highlighting the importance of being open and of adopting a constructive and humanitarian response to harm, rather than a defensive and legalistic approach (see online box 1 for an overview of existing principles and guidelines in the UK).19–25

In the USA, a national survey study of risk managers found that although the majority of hospitals reported disclosing harm to patients at least some of the time, only one-third of hospitals had board-approved policies in place.26 Respondents were much less likely to disclose preventable harms than to disclose non-preventable harms of comparable severity partly because of fear of litigation. Commonly quoted barriers to open disclosure include discomfort and a lack of training on how to disclose, a fear of litigation, a culture of infallibility among health professionals and inadequate systems for analysis, discussion and learning from mistakes.27 ,28 In Britain the National Patient Safety Agency (NPSA) released guidance on ‘Being Open’ in 2005 and then, concerned about the lack of impact, re-released the document in 2009 (note that the NPSA has now been disbanded).24 The Being Open framework provides best practice guidance on how to create an open and honest environment in which patients, their families and carers receive the information they need to understand what went wrong, and the reassurance that everything possible will be done to prevent recurrence; and all involved, patients, families and healthcare professionals, feel supported.

Currently the extent to which NHS organisations have endorsed such guidelines and put them into practice is unknown. We aimed to investigate the current state of practice in English NHS Trusts in relation to the communication of serious patient safety incidents to patients and families and the provision of support for patients, families and staff in their aftermath. Serious patient safety incidents were defined as incidents that result in unexpected or avoidable patient death, permanent, long-lasting harm, outcome that requires major life-saving intervention, prolonged pain or psychological harm, as per the NPSA's 2010 national framework for reporting and learning from serious incidents requiring investigation. Such incidents are the result of treatment or care rather than the result of a patient's illness or underlying condition.29


Study design

A cross-sectional quantitative design using online surveys and purposive sampling was used.

Setting and participants


In collaboration with the NHS Litigation Authority,30 we identified the principal risk or patient safety manager in all NHS Trusts in England. These managers would have responsibility for the implementation of the Being Open policy and also be involved in the management and follow-up of patients involved in serious incidents. They are therefore best placed to give an accurate presentation of the organisational policies, procedures and actions after serious incidents based on their personal impressions of how such policies are implemented at their Trusts. The respondents have a variety of formal titles but in this paper we refer to them as ‘patient safety managers’.

Procedure and materials

Participant recruitment

An invitation email for participation to the study was sent to one person who fulfilled the participant inclusion criteria in each English NHS Trust (n=386). The email included a description of the study and the link to the online survey. Reminder emails were sent to participants every 3 weeks. Data collection took place between November 2010 and February 2011.

Online survey

A questionnaire was developed through a comprehensive review of existing relevant literature2–5 ,8 ,15 ,17 ,28 ,31–40 and consultation with risk and patient safety managers, clinicians and experts in patient safety and quality improvement. The questionnaire included questions on hospital policies and practices relating to the management of the aftermath of serious patient safety incidents.

Specific scales were developed to explore the following areas:

  • (a) Availability of local policies on being open, frequency of being open and change of frequency of being open over the last 2 years.

  • (b) Structure (ie, parties involved in being open discussions, timeframe and regularity of being open discussions) and content of discussions with patients and their families.

  • (c) Barriers to being open with patients and families (see figure 1 for specific items).

  • (d) Availability of different forms of support for patients and families (see table 3 for specific items).

  • (e) Importance and availability of different forms of support for staff (see table 4 for specific items).

Items on the content of discussions with patients and families and the availability of support for patients and families in the aftermath of serious incidents were repeated for three different levels of incident severity: incidents leading to patient death, injuries leading to severe harm and incidents from which the patient is likely to make a full recovery. An open-ended question at the end of the survey prompted participants to provide general comments about being open in the aftermath of serious patient safety incidents.

The questionnaire was delivered using the ‘Survey Monkey’ online survey tool and was piloted with 10 risk managers who provided feedback for relevance of questions, comprehensibility, readability and length. Changes were made to the survey as appropriate based on feedback from the pilot participants and experts in patient safety research and quality improvement.

Data analysis

Data were analysed using SPSS for Windows V.18. Descriptive statistics were used to summarise participants’ ratings of patient and staff support policies and practices. Independent group t tests were employed to investigate differences in the uptake of Being Open policies between acute and non-acute Trusts. Repeated measures analyses of variance were used to investigate differences in the compliance of Trusts with openness and other patient support practices for three different levels of incident severity. Cases with missing data in each main section of the survey were omitted from the corresponding analyses.


Participant characteristics

We received partially or wholly completed surveys from 209 respondents, corresponding to a 54% response rate. A total of 153 participants (40%) completed the survey in its entirety. Table 1 shows the participant characteristics for the total sample (n=209).

Table 1

Participants’ characteristics

Awareness of Being Open and related policies

Almost all participants, 98% (204 of 209), reported that they are familiar with the Being Open guidance issued by the NPSA and 91% (190) of Trusts had a board-approved policy in place. Six years after the first release of the guidance, therefore, 9% of Trusts do not have a board policy on open disclosure. Less than half the managers 44% (91) reported that the guidance is followed all the time, with 38% (80) saying that it was followed more than half the time, and 18% less than half the time. Twenty-two per cent (46) of the respondents thought that the number of open discussions has not changed over the last 2 years in their Trusts, 38% (80) thought that it had increased slightly and another 38% (79) reported a substantial increase. We found no significant difference between the acute Trusts and community Trusts on either the extent of the guidance followed (mean=4.23, SD=0.90 vs mean=4.06, SD=1.04; t(205)=−1.28, p>0.05) or the change in use of the guidance over time (mean=4.19, SD=0.86 vs mean=4.04, SD=0.83; t(207)=−1.31, p>0.05).

Frequency of open discussions with patients/families by incident severity

Incidents are much more likely to be discussed openly with families when they are of a more serious nature (F(1.46, 22.11)=25.06, p<0.001). Pairwise comparisons confirmed that incidents that are followed by full recovery (mean=3.93, SD=1.06) are discussed openly significantly less often than incidents that lead to death (mean=4.31, SD=9.82) (t=−0.373, p<0.001) or incidents that lead to severe long-lasting disability (mean=4.27, SD=9.28) (t=−0.340, p<0.001).

Structure of Being Open meetings with patients/families

When open disclosure meetings do take place executives and managers are more strongly represented than clinicians, though the lead clinician on an investigation is often present. Table 2 shows the proportion of various people reported as being routinely involved in discussions with patients and families, the timeframe of such discussions and their regularity. About a third of open disclosure meetings take place within 3 weeks, but two-thirds take place 3–6 weeks after the investigation.

Table 2

Structure of ‘Being Open’ meetings for serious patient safety incidents

Content of discussions with patients and families by incident severity

Table 3 shows the relative frequency of a range of elements of open disclosure meetings. ‘Sharing investigation findings’ was reported as the most frequent component of such discussions while ‘taking responsibility for harm’ was the least frequent. All elements of the discussion are significantly less likely to occur for less serious incidents (F(1.29, 19.70)=12.47, p<0.001) with all elements being followed significantly less frequently for incidents from which patients recover fully than for incidents that lead to death or severe long-lasting disability.

Table 3

Frequency of different elements in open disclosure meetings with patients and families and support services for patients and families after serious patient safety incidents*

Frequency of support practices for patients and families by incident severity

Table 3 shows the frequency at which a range of support practices were reported to take place after discussions of serious patient safety incidents with patients and families. ‘Provision of details of a person of contact’ was the only common practice and this did not always occur. As before, these practices are followed less frequently for less serious incidents than for those that result in severe long-lasting disability (F(1.66, 253.22)=56.90, p<0.001).

Barriers to being open with patients and families

The most highly rated barriers to being open were a fear of negative patient and family reactions, fear of litigation and managers’ concerns about clinicians’ response to being accused of malpractice. Managers had fewer anxieties about confidentiality or anonymity and, encouragingly, did not fear a lack of institutional support for being open (see figure 1). Patient safety managers’ comments at the end of the survey highlighted a number of important barriers to them being as open as they would like, suggesting overall that while the policies may have been adopted in the majority of NHS organisations much more needs to be done to make their use a daily reality (see online box 2).

Figure 1

Perceived importance of barriers to being open with patients and families after serious patient safety incidents. The response scale was a five-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Importance and availability of support practices for staff

Patient safety managers considered that highly important (ie, mean score near or higher than 4) forms of support for staff were guidance by clinical supervisors, prompt debriefing, information about processes after serious patient safety incidents, peer support groups, occupational health and psychological services, exemption from clinical duties after serious incidents and training on being open; the least highly rated (mean score lower than 4) were chaplaincy support, stress management services and personal legal advice. In terms of availability, certain forms of support that were rated as highly important were not rated as always available (ie, guidance by clinical supervisors, prompt debriefing, exemption from duties after serious incidents, peer support groups and training on being open) (see table 4).

Table 4

Perceived importance and availability of support practices for staff who are involved in serious patient safety incidents*


This is the first study to our knowledge attempting to examine in detail the current state of practice in English NHS Trusts in relation to being open and supporting patients and healthcare professionals in the aftermath of serious patient safety incidents.

The study findings showed that almost all patient safety managers who completed the survey are familiar with the Being Open guidance.24 Nevertheless, 6 years after the initial release of the guidance 9% of Trusts do not have a policy in place and 18% of Trusts openly discuss incidents with patients and families half, or less than half, of the time. Disclosure policies seem to be implemented less frequently for less serious incidents, that is, those that do not lead to serious harm or death. A potential explanation for this could be that organisations are more likely to be open when there are higher expectations of a patient complaint. For instance, Bismark et al in New Zealand found that the propensity of injured patients to complain increased significantly with incident severity: odds of complaint were 11 times greater after serious permanent injuries than after temporary injuries, and 18 times greater after deaths.41 Mazor et al in a study of patients' attitudes to hypothetical scenarios of medical errors found that patients' intention to seek legal advice was significantly higher for non-disclosure in the case of a life-threatening outcome than in the case of a less serious outcome.42 Most patient safety managers reported that the frequency of such discussions is increasing, which potentially reflects the emphasis that being open has received recently at the policy level.

Being open with patients in the aftermath of patient safety incidents poses considerable challenges for the organisation and for the clinical staff involved. A host of different barriers have been suggested by empirical studies, such as discomfort and clinical staff's lack of training on how to disclose, a fear of litigation, a culture of infallibility among healthcare professionals and inadequate systems for analysis, discussion and learning from mistakes.27 ,28 Patient safety managers in this study considered that the most important barriers to open discussions with patients and families concern clinical staff's understandable fear of litigation, blame and the reaction of the patient and family. They tended to view institutional barriers as less problematic, in contrast to findings from other studies focusing on clinicians' views.4 ,43–45 Patient safety managers, with a more institutional perspective, perhaps see greater clinical barriers to disclosure whereas clinicians perceive more managerial and organisational barriers. The range of factors that this and previous studies have suggested as potential barriers to being open constitute the context within which guidelines such as the NPSA's Being Open framework are being implemented, therefore it would be valuable in future studies to understand the contextual factors that impede or facilitate openness from the different perspectives of those involved.

Informing patients about what happened and how is of huge importance in maintaining trust, as studies on patients' and the public's attitudes to medical incident disclosure also highlight.3 ,5 ,39 ,46 When meetings do occur there are signs that organisations commonly share findings, provide explanations and apologies but less frequently take responsibility for the harm caused. The causes of the incidents are no doubt complex and so may not be due to poor care, so organisations would not necessarily wish to shoulder responsibility even in serious cases and are probably also influenced by fear of potential litigation.4 ,31 However, accepting responsibility for harm has been associated with higher patient satisfaction,37 and our finding that this is followed less than half the time could be damaging for effective discussions with patients and families. A particularly worrying finding is that a high proportion of patient safety managers reported that the first open meeting with patients and families takes place 3–6 months after the investigation, which itself may have lasted some months. This implies that many patients and families are waiting up to a year before any clear explanation of serious incidents is given, which is extremely stressful in many cases. Patients are in favour of prompt provision of information about things that go wrong in their care,3 ,5 ,39 ,46 and this delay in itself is likely to exacerbate the distress already caused, erode trust and may also increase the likelihood of further complaints and litigation. The survey also suggests that there is limited representation of clinical staff in comparison to managers at open disclosure meetings, despite evidence suggesting that patients attach great importance to clinical staff who are involved in patient safety incidents being also involved in the communication of the incidents to the patients.9

The provision of support to patients and families appears to be much less well established than open disclosure. Provision of details of a contact person was commonly, but not always, offered to patients whereas other forms of support were offered on a ‘half the time or less than half the time’ basis. A qualitative study investigating patients’ experiences of actual disclosures highlighted that while the initial disclosure and explanation is critical a host of other issues may weigh heavily in patients’ minds in the aftermath of adverse events. Patients and families, for instance, want financial and other support, changes in practice or an opportunity for patients to meet with staff originally involved in the adverse event.39 More recently, patients and family members who were involved in high severity incidents in Australia confirmed that open disclosure still falls short of patient and family member expectations, such as their need for more follow-up support.47

Finally, support for staff was clearly recognised to be important in the aftermath of serious incidents but, equally clearly, did not seem to be available to the degree that it was thought to be important. Training on being open, especially, was recognised by our participants as an important structure for staff but was reported as available less than half the time. Provision of guidance for staff on open disclosure and support for staff is almost certainly needed on a wide scale, given that only a small proportion of incidents come to the attention of patient safety managers. For instance, recent qualitative work that we have conducted with surgeons on their experiences of patient complications indicates that surgeons still face a range of dilemmas pertinent to when, how and to what extent they can be open with patients when serious complications occur. Surgeons often feel conflicts between their human response, their professional obligations and their role as employees. The interests of the Trust can sometimes be the most predominant factor in how complications are communicated to patients and their families. Findings from this ongoing piece of work, in conjunction with existing literature,14 ,15 suggest that clinicians' ability to cope with the emotional impact of adverse events is very much dependent on available reassurance and opportunities for learning.

Limitations and challenges of evaluation of compliance with Being Open

The above findings need to be interpreted in the light of certain limitations. First, our participants hold management roles in their organisations and it is likely that their perceptions differ from those of frontline staff. Nevertheless, patient safety managers, as key informants of patient safety policies, are well placed to provide an authoritative view on organisational policy implementation. Another limitation of the study is the fact that the survey was distributed to only one person per organisation. However, our participants are the ones that were nominated by their Trusts as the most relevant for the provision of information on patient and staff support practices after serious patient safety incidents.

We also fully recognise that the use of an online survey is just a first step in a much needed exploration of open disclosure and support for patients and staff. Further work using mixed-methods approaches is needed to understand the impact on patients and staff and the wider organisation, and also to assess a range of incidents and responses to them. The views of patient safety managers are important in reflecting institutional policy and intentions but they may well differ considerably from the experiences of patients and staff involved in serious incidents. Also, unlike many other policies, the success of implementing Being Open cannot easily be determined by asking whether an explanation and apology has been offered. Following the policy and guidance may not always equal delivering open disclosure, as patients must actually feel that they have had an open and honest discussion and that they have been genuinely supported after the incident. Similarly, asking healthcare professionals' views on the challenges of being open as well as their preferences for support will help organisations to respond more effectively after serious incidents. Further qualitative or mixed methods research is needed to assess patients' and healthcare professionals' experiences after patient safety incidents to better understand the views and needs of those directly involved in them.


Our findings suggest that there is high awareness among patient safety managers of the importance of being more open with patients, but that progress is slow and that some Trusts have simply failed to recognise the importance of this issue. Our findings highlight the need for NHS Trusts to look closely into the ways in which they manage the aftermath of patient safety incidents and to ensure that sensitive support mechanisms are in place for patients, families and staff. More research is needed to clarify patients' needs and healthcare professionals' preferences for support when incidents occur to inform how best to support and help patients, families and staff in the aftermath of serious incidents.


We wish to thank the participants of this study. We also wish to thank Ruth Symmons, Gaynor Pickavance and Emma Corbett for their help with the piloting and the distribution of the survey.


Supplementary materials

  • Press release

    Press release

    Files in this Data Supplement:


  • Funding This work was supported by funding from the Health Foundation. The Imperial Centre for Patient Safety and Service Quality is funded by the UK National Institute for Health Research (NIHR).

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Correction
    BMJ Publishing Group Ltd