Informing best practice for conducting morbidity and mortality reviews: a literature review*

Preventable hospital mortality is a serious public health issue, particularly when mortalities are associated with events that are preventable. As a result, there is a clear need to have rigorous, systematic and effective processes in place to enable the assessment of quality of care in a manner that is timely. A mortality and morbidity review (MMR) is a peer-review process that aims to provide medical education, improve patient care and increase clinical performance in cases where morbidities and mortalities occur. MMRs provide an open, collaborative and transparent review process for clinicians to examine practice and identify areas of improvement, such as patient outcomes and adverse events, without fear of blame or individual focus.^1–3 Terms such as ‘morbidity and mortality conference’, ‘morbidity and mortality meeting’ and ‘patient safety and morbidity and mortality conferences’ are also used interchangeably throughout the literature.

Historically, MMRs are commonly conducted in surgical departments as a mode of clinical education and a way of reviewing and improving practice.^4–9 However, MMRs have been conducted in a variety of settings, including acute care units,^10,11 community medical centres,¹² emergency departments,¹³ general medical units,¹ intensive care units¹⁴ and palliative care units.¹⁵ In fact, standardised MMRs have been deployed and evaluated across entire hospitals or hospital networks.^16–19 However, the declared purpose of MMRs varies, with the most commonly reported goals being teaching,^5,7 medical management or quality improvement.^4–9 Some have suggested that an effective MMR should contain certain elements, such as: identification of events resulting in adverse patient outcomes; fostering discussion of those events; identification and dissemination of information and insights about patient care that are drawn from experience; reinforcing accountability for providing high-quality care; and creating a forum in which physicians acknowledge and address reasons for mistakes.²⁰

In Australia, health care providers independently organise and operate MMRs. There is no mandatory process or procedure that must occur when cases meet the requirements of a review. Instead, organisations conduct MMRs internally under their own direction. As a result, there is an absence of a consistent, standardised approach and it is likely that MMRs differ from healthcare network sites, departments and units. Although MMRs may be commonplace in a range of healthcare settings, the characteristics of how they are conducted, in terms of structure and format, vary considerably.^1,8,21 A strength of the MMR process is that a structured methodology can be applied to each and every case with the view of systematically, and transparently, examining cases without bias or predisposition and with a view to system and process improvement.¹⁸ After exploring the information regarding preventable death in our network (Monash Health), it became apparent that there was a lack of consistency in how MMRs were being conducted. Therefore, we were interested in exploring the literature to inform the development of a gold standard or best practice. We were also interested in describing the outcomes and common features of MMRs.

A search for peer-reviewed and grey literature published from 2009 to February 2016 was conducted in PubMed and Google. This time period was selected because the present review updated a previous review conducted on MMRs from inception until 2009.²⁰ Documents identified were screened using inclusion and exclusion criteria established a priori and based on the Population-Intervention-Comparison-Outcome (PICO) framework (Table 1).²¹ Only documents that covered the mortality review process in a hospital setting were included, but items that included ‘mortality’ and ‘morbidity’ were not excluded.

Table 1.  Document eligibility criteria
MMR, morbidity and mortality review; MMC, morbidity and mortality conference; MMM, morbidity and mortality meeting; PSMMC, patient safety and morbidity and mortality conferences

Specific search terms selected were generated from a combination of those used in previous MMR reviews²⁰ and searches of appropriate MeSH terms (Table 2). Reference lists and citing articles were also screened for potential inclusions.

Table 2.  Search terms used to identify documents in PubMed and using Google

There are several terms that refer to the process of formally discussing cases of morbidity and mortality, including ‘morbidity and mortality review’, ‘morbidity and mortality conference’ and ‘morbidity and mortality meeting’. To simplify, we refer to these clinical meetings as MMRs from this point onwards.

PubMed search results were exported into Endnote (X7; Thomson Reuters), where titles and abstracts were screened and duplicates were removed by one reviewer (CWJ). All potential papers were searched for inclusion and the full text was obtained where required. Following the Google search for grey literature, filters for English language and date were applied, and all entries were screened for inclusion by two reviewers (CWJ, MG). Those links that were relevant were then exported to a Microsoft Word document.

For each study identified from the PubMed search, details regarding setting, study design, reported outcomes, tool type, clinicians present and the timing of MMRs were extracted (Table 3). Similarly, data regarding setting, tool type, clinicians present and timing were also extracted for the grey literature items (Table 4).

Table 3.  Comparison of morbidity and mortality review (MMR) characteristics for each paper
MDT, multidisciplinary team; ICU, intensive care unit; CMO, chief medical officer; QI, quality improvement; RN, registered nurse; NHS, National Health Service; MMC, morbidity and mortality conference; SBAR, Situation, Background, Assessment and Analysis, Recommendation; MI, myocardial infarction; MMPRC, morbidity and mortality peer review conferences; VAP, ventilator-associated pneumonia

Table 4.  Comparison of morbidity and mortality review (MMR) characteristics for grey literature items
MDT, multidisciplinary team; SBAR, Situation, Background, Assessment and Analysis, Recommendation; *, unable to be determined

The PubMed search identified 827 publications, of which 27 full-text articles were retrieved (Table 3). The search of the grey literature identified 256 results. After screening, 31 documents were included in the present review (20 peer reviewed articles and 11 items of grey literature) (Fig. 1).^{1–19,22–33} After extracting and summarising the relevant information, the evidence presented herein relates to specific outcomes examined and reported in the literature, as well as common features of MMRs.

Fig. 1.  Search results and screening process used in the present review.

Five studies explored mortality rates as an outcome.^4,7,10,11,23 One study evaluated the effect of MMR on learning.⁵ Four studies explored satisfaction levels of how MMRs were conducted by participants.^12,17,18,24 Six studies described or evaluated the use of an MMR tool.^{1,6,8,9,14,16} One study explored cost,¹² two studies explored the format of MMRs^13,19 and four studies explored quality of care or care delivery.^1,15,22,23 The validity and reliability of a structured MMR tool has also been reported.⁹

Six studies used MMRs in a surgical setting,^4–9 with other settings including acute care,^10,11 community,¹² emergency care,¹³ intensive care,¹⁴ palliative care,¹⁵ general medical¹ and paediatric anaesthesia.²⁴ Nearly all items of evidence reported information about the timing of their MMRs.^{1–6,8,9,11–15,17–19,24,25} Fifteen items of evidence used multidisciplinary teams in their MMRs;^{1–3,5,6,11,14–17,19,22–25,28} others included physicians,^10,13,18 surgeons^4,7,8 and nurses.¹²

The present systematic review was conducted to update a previous review,²⁰ and identified and synthesised recent evidence about MMRs. There is limited evidence in the literature evaluating outcomes or measurements of patient-centred care as a result of MMRs. A previous review has described other literature on the effectiveness of specific MMR approaches and identified MMRs to be effective in identifying and engaging clinicians in system improvements, reducing deaths from cholera and creating a safe forum for discussion of errors for junior medical staff, including removing fear of incrimination.²⁰ Mortality rates, satisfaction, cost, quality of care and patient safety were reported in the studies included in the present review.

MMRs provide a quality management approach that can improve mortality rates in hospitals with sub-optimal performance.¹⁰ It should be noted that the finding of mortality rate improvement is based on observational, non-randomised data. Gross mortality and age-adjusted mortality are reduced following the implementation of MMRs.^4,23 Structured MMR proforma capture more mortalities, morbidities and adverse events than standard MMRs.⁷ In addition, MMRs decrease medical malpractice claim rates.¹¹

MMRs lead to higher levels of clinician satisfaction because they provide rigorous case discussion, collaboration, understanding of operations of non-clinical staff and communication between all staff.¹⁸ The provision of a clear framework means that meetings are well organised, and the educational tone leads to a better appreciation of the topic.²⁴ However, time restrictions mean that the discussion of cases and topics is limited.²⁴ Neurosurgeons, anaesthetists and head nurses are satisfied with MMRs because they provide useful solutions to clinical practice problems.⁶ MMRs have a positive effect on teamwork, offering opportunities to discuss deficiencies and to decrease tensions.¹⁹ It is also suggested that MMRs promote a safety culture through the identification, analysis and correction of deficiencies.¹⁹ Alternately, there is some dissatisfaction in MMRs that contain root cause analysis because there are issues in training staff in this technique.¹⁸

Cost savings are evident following the implementation of an MMR.¹² In patients with ventilator-associated pneumonia, MMRs have been shown to reduce ventilator-associated pneumonia cases leading to a cost saving of US$13 333 per month.¹²

MMRs identify quality issues more effectively.⁴ Patient care is improved in palliative care environments with pain, dyspnoea, nausea and agitation all addressed at a higher rate.¹⁵ MMRs lead to improvements in patient care via enhanced governance and clinical management follow-up.¹

In addition, those that lead MMRs believe patient care is enhanced through improved teamwork and an improved platform to discuss deficiencies and decrease tensions.¹⁹ MMRs are a powerful driver of the safety culture, with increases in motivation, resulting in improvements in harm minimisation practice and improved promotion of organisational learning.²³ Information and data from MMRs are positively implemented into clinical practice via educational programs to allow for standardisation of best practice.²³ The ability of MMRs to engage multiple members of the healthcare team in a discussion of adverse outcomes while collaboratively focusing on solving problems also leads to improvements in patient safety.²² This is brought about by the identification of potential system failures, empowering workgroups to address specific systems-based problems and making transparent accountability for regular follow-up.²²

To build a picture of what the ideal MMR looks like, the most common features were identified from the literature, as summarised below.

There is large variability and no clear consensus on the ideal timing of when an MMR should occur. The number of reported cases can often determine the frequency of meetings.¹⁹ Options used throughout the literature are within 1 week of an event, whenever practical within days of the event, within 24 h of an event, weekly,^{1,4,11,12,15,17,18} monthly,^{1,5,8,9,11,13,14,22,24} bimonthly¹¹ and 3 monthly.⁶ There has not been empirical evaluation of timing.

Evidence regarding meeting duration varies, with MMRs going from 20 min up to 4 h.^1,20 However, a 15-min presentation with 5 min for questions has been used previously.^8,17 In addition, 30 min per case has been used.¹⁸

Leaders should have high skills and expertise in the area of morbidity and mortality cases.² Leaders should be trained or have skills in auditing, the ability to understand and interpret the clinical information accurately, the ability to access senior medical advice and have an understanding of the clinical environment.² Leaders may include consultants, physicians or senior doctors or nurses; however, this role could be performed by any suitably trained staff member with access to senior medical advice as required.

Most of the literature reports the use of a multidisciplinary team.^{1,5,6,11,14–17,19,22,24} Participants may include any of the providers involved in the care of the patient, selected experts and others who can contribute to the analysis of the event and to the development of practical recommendations to improve patient safety.²⁵ The vast array of staff involved in MMRs is listed in Tables 3 and 4.

Case presentations vary in structure throughout the literature. The Situation, Background, Assessment and Analysis, Recommendation (SBAR) format is perhaps the most structured and best evaluated format.¹⁷ Some have suggested that in the month leading up to the MMR a core team should meet to gather, review and summarise information from the patient’s hospitalisation in a time series flow diagram.²² A brief literature review of the disease or illness specific to the case is presented.²² Other reviews are formally prepared in presentation format and shared with those attending the MMR 2 days before the meeting.⁴ Cause-and-effect diagrams have been used to identify factors that may have been related or contributed to the adverse outcome,²² as well as root cause analysis to identify quality issues and causes of error.^4,17 In addition, a seven-step process has been used outlining case identification, case classification, case preparation and review, case analysis, case discussion, recommendations and closure or follow-up.¹⁴

Clear and defined terminology when describing MMRs is important.¹ Only factual information in incident or occurrence reports is provided, and restraint needs to be applied to statements of blame, speculation, opinion or other commentary as to the reasons for what happened.^9,17,19,25 Emphasis must be placed on MMRs being a safe, supportive and blame-free forum to facilitate improvement and accountability.

The present review found limited evidence regarding agenda. The following outline has been used previously:²² (1) a reminder of the systems-based approach and confidentiality by the MMR leader (5 min); (2) a review of task force progress from prior conferences by the MMR team (10 min); (3) a case presentation in a timeline format by resident leaders (10 min); (4) a brief literature review relevant to the case in question by resident leaders (5 min); (5) identification of key issues leading to the undesired outcome by all attending the MMR (25 min); (6) identification of workgroups to address the key issues by the MMR team (10 min); (7) a reminder of confidentiality by the MMR leader (5 min); and (8) evaluation of the conference by those who attended as administered by the MMR leader (5 min).

There is varied evidence for positions responsible for follow-up within an organisation. Leadership can be involved in the findings of MMRs.^2,25,29 In addition, risk or quality teams are involved in the process because findings may be the result of issues with system processes.^1,2,25 Feedback functions have also been built in to online systems that automatically administer MMR forms where necessary and track the process of each MMR case.¹¹ To assist with evidence translation and process improvement, improvement measures determined by the MMRs should also be communicated to frontline staff through ward-based speciality governance and clinical management positions or staff.¹

A previous review identified that:

Follow up tends to be limited to either case summary reports and or designating individuals to follow up actions…some studies have reported more comprehensive approaches including documentation of outcomes, evaluations, development of action plans, verbal updates at subsequent MMRs, written reports and tracking of actions.²⁰

The present review is not without limitations. Only one database was searched for relevant literature and one author reviewed the references retrieved. To overcome these limitations, the authors used secondary searching of citations and bibliographies to maximise the capture of studies that did not show up in the database search or were inappropriately excluded.

The result of the present review has since informed and steered a large organisational change. As such, the evidence in the present review has been used to frame a gap analysis between best practice and the current organisational procedures for MMR. Subsequently, items of best practice have been prioritised and reporting structures improved with terms of reference, procedures and implementation tools developed. These changes are serving as a benchmark that the organisation has selected to measure, monitor and gain feedback at the executive level in order to inform policy development and improve the quality of patient care in the future. More widely, the present review informs clinical practice by providing a current summary of evidence around MMRs.

There is very limited peer-reviewed literature exploring patient-centred outcomes as a result of the MMR process. MMRs have resulted in a decrease in gross mortality and are effective in identifying and engaging clinicians in system improvements, reducing deaths from cholera and creating a safe forum for the discussion of errors for junior medical staff, including removing fear of incrimination. Based on the best available evidence we are able to inform best practice in conducting MMRs. This includes an MMR that should not focus on the actions of individuals, rather on education and/or quality improvement. MMRs should include an agenda, a structured presentation format (i.e. SBAR), an analysis of error process and should conclude with actions to be performed, and these followed-up at the beginning of subsequent MMRs. MMRs should consist of a multidisciplinary team, including those who had the most contact with the patient. Finally, given the features are general in nature, it is possible for a single standardised MMR to be implemented hospital wide.

None declared.