Article Text
PDF
Statistics from Altmetric.com
Introduction
Although morbidity and mortality (M&M) meetings are regarded as the cornerstone of hospital governance process,1 there is little evidence that they are effective in improving patient outcomes.2 ,3 Formats vary widely, approaches to categorising error inconsistent4 ,5 and few meetings adopt accepted models for incident analysis.6 Many errors are not reviewed, and the key protagonists often not present when a case is being discussed. While a standardised approach may engage the healthcare team in quality improvement initiatives, such formats have not been shown to impact on specific patient outcomes.7 Furthermore, M&M meetings fail to engage affected families. This lack of transparency in the context of the Francis report is at odds with our duty of candour to patients when things go wrong.8 There is, therefore, a need to re-evaluate how we should best scrutinise morbidity and mortality to bring about tangible improvements in patient care.9
The initial challenge is one of definition. The term ‘mortality’ is universally understood and describes a sentinel event that is easily recorded. However, professionals interpret ‘morbidity’ in a variety of ways, which partly explains why M&M meetings often fail to assess non-lethal patient outcomes in a systematic fashion. For the purposes of this article, I have taken ‘morbidity’ to encompass two situations: medical error defined as the failure of a planned action to be completed as intended, and adverse events as unintended patient harm resulting from either healthcare intervention or omission. Adverse events may occur due to medical error, but are often the unavoidable consequence of medical treatment, that may or may not be predictable (figure 1).10
Schema illustrating medical error and adverse events in the context of treatment, disease and injury.
In this paper, I will suggest that the investigation of medical error, adverse events and child death each require a distinctive approach. I will then discuss the contentious issue of ‘avoidability’, before concluding with an analysis of how proactive risk management should occur within a framework of coordinated clinical governance that achieves completion of a safety feedback cycle.
Risk management
Medical error
An effective clinical governance programme should identify and scrutinise medical errors that are both harmful and non-harmful to patients. Medical error may be active (occurring at the level of the healthcare practitioner) or latent (occurring at the level of the organisation). The consequences of latent errors are typically dormant until certain circumstances arise.11 Most medical errors do not result in injury to the patient. In healthcare, Vincent has proposed seven factors that contribute to medical errors: patient factors, task factors, individual factors, team factors, working conditions, organisational factors, and institutional context.12
Hospital departments usually record medical error through a process of ‘incident reporting’. High-risk specialities such as intensive care generate a large number of incident reports. In the USA, the Society of Critical Care Medicine has developed a voluntary, anonymous, web-based Intensive Care Unit Safety Reporting System (ICUSRS) to capture patient safety incidents across adult and paediatric ICUs.13 Drug errors are a common cause of medical error in children. Although most cause no harm, serious injury may occasionally result.14 ,15 The review of drug errors requires assistance from a pharmacist to determine whether the error occurred during the ordering, transcribing, dispensing or administration of the drug.16 The solution to drug errors is both at a department level, through pharmacist oversight and standardisation of processes, and at an organisational level, through investment in advance technologies.17 ,18 The clinical team must ensure that all incident reports are reviewed in a time-sensitive fashion. This is best performed by a designated clinician and nurse who may either conclude that the incident can be ‘closed’ or that it requires immediate action. Since many incidents are uncommon, it is also necessary to systematically review them in a themed fashion at a departmental clinical governance meeting.
Adverse events (ie, patient harm)
Adverse events may result from medical error but also from necessary medical or surgical treatment. An example of the former might be a child who develops acute kidney injury from iatrogenic aminoglycoside toxicity. Example of the latter might be a child left with reduced fertility after radiotherapy or a hemiplegia following neurosurgery for a brain tumour. Adverse events may result in permanent injury or death to the patient, and those that occur as a result of medical error can result in costly litigation.19–21 While it is acknowledged that adverse events are common, methodological heterogeneity between studies limits our ability to quantify prevalence across different healthcare settings. In critical care, patients who die have higher rates of adverse events than those who survive.22 In prehospital care, adverse events, such as the failure to recognise severity of illness, are common in children who subsequently die in a paediatric intensive care unit.23
The approach used to investigate an adverse event associated with medical error differs from one associated with necessary medical or surgical treatment. Patient harm associated with medical error lends itself to incident reporting in the same fashion as medical error generally. In the author's hospital, such events trigger a ‘rapid-response’ meeting in which three matters are determined: (a) Are there other immediate safety concerns that may affect the care of other patients? (b) Are the needs of the family being addressed? (c) Will a root cause analysis (RCA) investigation add value to understanding the event, and what should be the terms of reference of that investigation? A ‘rapid-response’ meeting occurs within 24 h and involves the clinical director, child's consultant and a representative from the patient safety team. The patient safety team leads the RCA investigation, which is attended by the key professionals involved in the incident.
Patient harm arising from necessary medical or surgical treatment may be unavoidable, but may reflect the skill of the treating doctor or the performance of the wider multidisciplinary team. In these situations, incident reporting has been shown to be a poor screening tool that, at best, flags potential areas of concern.24 Such events demand methodical clinician-led case note review that forms part of an overarching assessment of a team's performance. Therefore, the neurosurgical team's incidence of hemiplegia after tumour resection might be benchmarked against national and international peer data. This type of review lends itself to a regular clinical meeting, whose output informs a wider programme of clinical effectiveness that includes audit and research.
Mortality
Both medical errors and adverse events may occur in the context of a child's death and, if so, they should be recorded and investigated in the manner described above. However, a child's death also provides a window on the wider healthcare system. Since 1 April 2008, legislation in England and Wales has set out how child deaths should be reviewed.25 A prospective investigation of unexpected deaths and a retrospective review of all deaths by a child death overview panel provide a comprehensive dataset to complement national death registration statistics. Although this approach has been successfully adopted for reviewing unexpected deaths in the community, anecdotal experience suggests that a similarly systematic approach to hospital deaths does not occur.26 ,27
Conventionally, the M&M meeting only includes medical professionals in the hospital where the child died and does not engage individuals across the patient pathway. Let us consider a 3-year-old child who dies in the paediatric intensive care unit with meningococcal septicaemia. By design, the following professionals should be present at this child's death review meeting: the general practitioner who saw the child on the day before his collapse, the paramedic who attended his home, the local hospital medical team, the regional retrieval service and the clinical staff in the tertiary centre (figure 2). The timing of this meeting should be coordinated around the availability of these ‘key’ professionals, and videoconferencing and conference call technology facilitate this. This approach demands that a single death is reviewed at each meeting rather than several patients being discussed, which is what habitually occurs.
Key professionals present at an illustrative child death review meeting. GP, general practitioner; ITU, intensive treatment unit; PICU, paediatric intensive care unit.
We know from the Confidential Enquiry into Maternal and Child Health ‘Why Children Die’ study that, in a high proportion of cases, there are modifiable factors in the chain of events that lead to the child's death.28 M&M meetings often fail to go beyond identifying what the child died from, to considering why a child died of that condition at a particular place at a particular time. An informed analysis of child death requires clinical teams to move away from a simple medical model of death to one that recognises the bioecological reality of contributory factors. The Department of Education's child death B and C forms allow contributory factors to be recorded across domains intrinsic to the child, to family and environment, to parenting capacity and to service delivery.29 A preventable death is defined as one having modifiable factors, which, by means of a locally or nationally achieved intervention, might reduce the risk of future deaths. At a national level, such standardised metrics will enrich our understanding of the causes of child death and improve our ability to prioritise healthcare interventions.
The family should also be provided with opportunities to be involved in the review. However, while there are calls to have parents present in such meetings, I maintain it remains essential to convene a specific meeting for professionals. The reasons for this are threefold: necessary discussion of ‘parenting capacity’ would be challenging; contextual reference to similar cases would breach patient confidentiality and clinicians should be able to discuss system failures in a manner free of retribution. These caveats aside, it is still possible to bring parental questions to the meeting as a standing agenda item, and there should be automatic arrangements to meet with the parents prior to and after the review.
A child death review (CDR) meeting should not take place until the RCA and postmortem have been reported. The former may provide intelligence relating to factors in service delivery, while the latter may provide vital information relating to genetic factors and unexpected pathology. Good practice dictates that the CDR meeting should generally occur within 3 months of the child's death, and will usually conclude matters except in deaths subject to coronial or criminal investigation.
A properly formatted CDR meeting should therefore encompass the following features (figure 3).
Features underpinning a properly formatted child death review meeting. MCCD, medical certificate cause of death; SMART, Specific, Measurable, Achievable, Realistic, Timely.
Avoidability
Much energy is spent in the NHS concluding whether errors, adverse incidents and deaths are ‘avoidable’ or ‘preventable’.30 These are subjective distinctions, and varied definitions make comparisons problematic. Attempts by organisations to determine ‘preventability’ often impede a culture of open reporting. Healthcare teams worry that concluding an adverse event is ‘preventable’ implies fault that may inevitably lead down a path of complaint and litigation. Rather than trying to determine whether or not an incident is preventable, healthcare organisations might adopt a more pragmatic approach—one that encourages clinical teams to instead focus their efforts on accurately determining whether medical error or patient harm has occurred, reviewing the contributory factors and ensuring that learning arises from the event. In paediatric cardiology, a culture of open reporting has contributed to understanding risk factors for complications that have improved outcomes for the population as a whole.31 ,32 What matters is not an arbitrary conclusion of ‘avoidability’, but professional analysis of the care that the child received. Such scrutiny reassures parents that we hold their child's life in the highest regard and may provide some comfort that their experience will benefit the treatment of other children.
Proactive risk management
Clinical governance is defined as ‘a framework through which NHS organizations continually improve the quality of their services and safeguard standards of care’. It encompasses risk management, clinical audit, education, patient experience and clinical effectiveness (DH, England, 1998). Risk management processes include the review of medical error, adverse events and child death; each of which lend itself to individual scrutiny in the manner described above.
However, if risk management is our vehicle for reducing patient harm, safety experts should reflect that, in ever more complex health environments, things go right far more often than they go wrong. The linear model of RCA, which tracks the origins of an adverse outcome back to fixed causal determinable events, is too simplistic. In contrast to traditional ‘Safety-I’ thinking, advocates of a ‘Safety-II’ approach regard the constant variation and adjustment in human performance, as the essential basis for flexibility that allow systems to perform well 99% of the time.33 In terms of incident investigation, the goal then becomes one of understanding how things usually go right, and using this knowledge as the platform for then explaining why they occasionally go wrong. In practice, Safety-I and Safety-II are complimentary. Clinicians still need to focus on things that go wrong. However, as figure 4 demonstrates, rare serious untoward incidents, although easy to identify, often have complex aetiology and are difficult to modify. In contrast, normative behaviour may be easier to understand and to influence.
Safety focus in relation to event probability (adapted from Hollnagel et al)33.
Coordinated clinical governance: achieving completion of a safety feedback cycle
Experience from high-risk industries teaches us that the credibility of governance programmes arises through demonstrating completion of a safety feedback cycle; that is to say, reporting and investigation must result in timely evidential actions that effectively address system vulnerabilities.34 Change management requires certain basic activities: the occurrence needs to be recorded, intelligence elicited from relevant staff, a structured framework used to investigate contributing factors, agreement of necessary actions, assignment of actions, feedback to relevant parties, evidential completion of actions and subsequent audit. Such activity should occur continuously ‘behind the scenes’. Actions need to be SMART (Specific, Measurable, Achievable, Realistic, Timely) and feed into the wider governance programme of the department and the hospital.35 A standardised approach to ensuring actions are properly recorded and assigned, promotes organisational memory of recurring themes.36 The departmental Clinical Governance meeting should monitor progress on actions arising though incident and CDRs, and importantly triangulate investigations relating to the same child.
Figure 5 presents a schema illustrating how the recording and investigation of error, adverse events, child death and positive behavioural performance might each have consequences that are surveyed at a departmental clinical governance meeting—the latter providing the essential vehicle for linking risk management issues with audit, clinical effectiveness and patient experience.
Coordinated clinical governance. RCA, root cause analysis.
Moving forward
Healthcare organisations are responsible for creating a culture, where patient safety is central to operational strategy: one that encourages clinical leadership, invests in credible data and ensures effective channels of communication. The NHS should reward organisations that foster a culture of openness rather than overzealous accountability. Not all mistakes in medicine cause measurable harm to patients, and not all harm to patients arises from preventable error. The traditional M&M meeting does not lend itself to the modern day patient safety agenda. It is reactive, rather than proactive, and does not focus on promoting learning environments and anticipating unexpected events.37 The investigation of medical error, adverse events and child mortality each requires a distinct approach that revolves around a continuous cycle of reporting, professional scrutiny and follow-through of SMART actions. These processes should separately feed into a properly formatted clinical governance meeting, the purpose of which is to provide assurance to hospital boards and other regulatory bodies that there exists coordinated oversight of risk management, clinical effectiveness, audit and patient experience. Moving forward, real resilience in our healthcare systems may require a paradigm shift in our understanding of the variability of behaviour that allows humans to continually adjust to competing pressures in the workplace environment.
In general, clinical governance needs to show increased ability to move from reflective learning to service transformation. A multilevel error-reduction strategy at every stage of the patient pathway can make a real impact.38 While many actions pertain to the local team, some require a response at a regional or national level. The latter are often the most important. Clinicians should not shy away from trying to effect change in the interest of their patients. Establishing ‘fit for purpose’ governance arrangements is the basis for such endeavour.
Acknowledgments
The author would like to thank Professor John Henderson for his advice on the manuscript.
References
Footnotes
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.