Problem

In 2019, Mayo Clinic announced a joint venture with Abu Dhabi Health Services Company (SEHA) in a collaborative development of their flagship hospital, Sheikh Shakhbout Medical City (SSMC), the largest healthcare network in the United Arab Emirates (UAE).1 The joint venture seeks to combine strengths of both Mayo Clinic’s and SSMC’s care systems to accelerate the evolution of SSMC into a high-quality tertiary academic medical centre. Abu Dhabi is the capital of the UAE and the largest of the seven emirates. In 2007, the Abu Dhabi government created two organisations: the Department of Health and SEHA.2 The Department of Health defines healthcare strategy and goals, whereas SEHA is responsible for operating the publicly funded health system of Abu Dhabi. SEHA (the Arabic word for ‘health’) is the largest healthcare network in the UAE and is responsible for public healthcare centres and hospital quality and safety standards.1 2 Since its formation, SSMC has strived to follow the quality and clinical risk recommendations defined by SEHA, namely, a review of all sentinel events and all deaths within 45 days. These reviews were facilitated through departmental morbidity and mortality (M&M) conferences, as well as focused peer reviews. SSMC determined that, to accelerate its progress toward high reliability and an improved safety culture, it needed to enhance its reviews by instituting the Mayo Clinic Care Review Process (MC CRP) for deaths.

During its early operations, SSMC identified that its review process, as originally planned, did not optimally suit its needs, primarily regarding timeliness. One key performance indicator (KPI) reported at a system level is the percentage of all deaths reviewed within 45 days. This average KPI was 78.3% in quarter 3 of 2020, 16.7 percentage points below the target of 95%. Factors contributing to this gap included that only select cases were reviewed, with an inconsistent selection process; clinician time to review cases was limited; and structured support for efficient and timely reviews was lacking. Another concern was inconsistency in the quality of organisational learning yielded by the reviews. The reviews did not elicit real opportunities for systematic improvement. An internal analysis suggested several reasons for this: (1) each department reviewed its own cases, and findings were not easily shared across departments; (2) reviews were completed only by physicians, without input from nurses and other clinicians; (3) reviews were often focused on peer review/personal performance rather than on systems/processes3; (4) tracking and analysis of improvement opportunities or dissemination of lessons learnt was limited; (5) there were few actionable insights for performance improvement activities; and (6) when improvement opportunities were identified, a consistent ‘closing the loop’ process did not exist to ensure that the improvements were achieved and sustained.

Given these factors, SSMC requested that Mayo Clinic help establish a care review process for deaths at SSMC, a process that has evolved beyond that described by Huddleston et al,4 Learning from Every Death. Thus, a dissemination project was specifically designed to help accelerate the establishment of the MC CRP, adapted for the local care environment.5 The project was a collaboration between the Mayo Clinic Global Quality Consulting team and the SSMC clinical and quality teams, led by the quality director and chief medical officers of SSMC. The project had to be accomplished exclusively via virtual contact because of travel restrictions imposed by the COVID-19 pandemic. This article describes how Mayo Clinic translated its care review process internationally.

This international engagement included the following key objectives:

• Assist in disseminating the MC CRP to SSMC, adjusting the process with consideration of SSMC’s organisational culture and current processes.

• Transfer knowledge about the tactics and methods to perform an in-depth review of the patient’s clinical course in an efficient manner.

• Coach the local team to focus on system/process-related issues or clinical decision-making as contributing factors to a serious adverse outcome or death.

• Teach the local teams to identify care issues that may not have contributed to the patient’s death but caused or had potential to cause harm or deviation from the standard of care.

• Increase the proportion of deaths reviewed within 45 days to meet the KPI target of 95%.

• Increase the number of opportunities for improvement, communication of these findings and development of improvement projects based on these findings.

Background

Healthcare organisations have dedicated substantial resources during the past two decades to bridge the quality and safety gaps described in the two sentinel works published by the US Institute of Medicine, ‘Crossing the Quality Chasm’6 and ‘To Err is Human’.7 These reports outlined a framework to evaluate healthcare quality and highlighted the high frequency of medical errors, preventable harm and death in healthcare settings, thus launching a patient safety movement. In the wake of these reports, much progress has occurred. Some estimates suggest that in the initial 15 years after these reports were published, related efforts saved more than 1.3 million lives.8 Despite such improvements, substantial opportunity remains to reduce harm and death that occurs during healthcare. To continue to accelerate the progress of patient safety efforts and move toward high reliability, emphasis has focused on the importance of building a patient-centric safety culture and enhancing an organisation’s learning system.9 Among others, key components of an effective learning system include (1) transparency (an environment in which adverse events or missed opportunities may be discussed in a non-punitive way with improvement as a goal), (2) convergence on standardised care so that ‘what is supposed to happen’ is agreed on by those delivering care and (3) focused efforts to detect opportunities for improvement that are well resourced to implement real change.

Before 2014, Mayo Clinic initiated a systematic review of the care delivered to each patient who died in our hospitals.4 As Mayo Clinic continued this review process, challenges were noted regarding the timeliness of reviews and feedback to the practice, which led to subsequent refinement of the review process, now called the MC CRP (online supplemental table). The most relevant process change was the addition of a weekly mortality review huddle for early closure of cases with no substantial opportunity for improvement. Thus, communication of minor opportunities was provided more quickly to the practice. This change also allowed for comprehensive review of cases that had greater opportunities for improvement. Additional benefits of the change were a decrease in the turnaround time between clinical events and review of those events and enhanced recall, accuracy and engagement.

Although the MC CRP continues to evolve, it has several key components. First, every case is reviewed by at least two clinicians—one physician and one nurse—by using a standardised taxonomy to characterise opportunities to improve care from admission until death. Second, findings are discussed with those involved soon after the care was delivered to improve accuracy of understanding and create the best reflections for learning. Third, opportunities to improve are brought forward to clinical teams to address in a timely fashion.

Over time, conducting systematic care reviews with the MC CRP has produced several benefits:

1. Increased situational awareness. The MC CRP revealed many new instances of adverse events that were not detected by using the Agency for Healthcare Research and Quality—defined patient safety indicators, clinician-reported events or the Institute for Healthcare Improvement Global Trigger Tool with physician confirmation.10

2. Strengthened team-based care. Discussion forums (huddles and committee review) offered a much-needed opportunity for clinicians to engage in open discussion in a multidisciplinary setting about standards of care and areas to improve delivery systems or policies. The huddle process occurred when staff still had recollection of events surrounding the care of the patient, which allowed for a perspective not possible with chart reviews conducted weeks later. Similar findings were reported in a study of an electronic review tool that was sent to frontline clinicians after inpatient deaths for real-time feedback on quality and safety issues.11

3. Accelerated learning cycles. Huddles promoted prompt review of cases, which allowed for closure of those with no opportunity for improvement and further comprehensive review of cases that reflected opportunities. More complicated reviews received more attention and had greater impact.

4. Enhanced engagement in improvement. Early engagement of clinicians throughout the review process was more satisfactory and led to increased involvement in process improvement.

5. Strengthened safety culture. The final and most transformative benefit was that the process strengthened the patient safety culture, which changed discussions surrounding adverse events from ‘who did what’ to ‘what happened and why’.

Our ‘mortality review’ process has evolved to a redefined ‘care review’ process that not only identifies system/process failures that may have contributed to the patient’s death, but also provides opportunities to improve the care that was delivered. Furthermore, a more streamlined approach to reviews has allowed the process to be less resource intensive, improved use of the taxonomy to enhance our understanding of which aspects of care must be improved and shortened the time for ‘closing the loop’ between discovery and improvement activities. Our CRP is supported by the use of an external, electronic database to facilitate the collection of reviews, track our opportunities for improvement and communicate findings to the clinical practice. The MC CRP differs from a typical peer review or M&M review with this taxonomy in that it addresses processes that can be improved instead of focusing on the caregivers. The ability to aggregate learnings at both site and institution levels allows for translation of individual case lessons to themes for system improvement. As a result, the organisation transforms into a learning system operationalising high reliability (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Multiple reviews build shared learning. The Mayo Clinic Care Review Process (MC CRP) is one of multiple reviews that can be completed for a patient encounter. It is part of a system of reviews that leads to shared learning. Shared learning occurs at the individual hospital level and also at the healthcare institution level.

Measurement

One of the main objective measures to evaluate the success of this project was the percentage of deaths reviewed within 45 days. The percentage of deaths reviewed was determined from the number of days elapsed between death and submission of a report containing the clinical and quality review to SEHA. During the three quarters before implementation of the new review process, death reviews were completed within 45 days for 62.7% of deaths in quarter 1, 2020, 88.0% in quarter 2 and 78.3% in quarter 3 (figure 2). These did not meet the KPI of 95% completion required by SEHA. With the quarter 3 value of 78.3% as the baseline measurement, this value was 16.7 percentage points below the target of 95%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Percentage of deaths reviewed within 45 days at Sheikh Shakhbout Medical City by quarter.

Before the engagement with Mayo Clinic, SSMC did not use a huddle process for reviews. This process was described and implemented in quarter 4 of 2020, and measurement of case closures in huddles were monitored starting at the same time.

In addition to this objective measure, another goal of the project was to engage more of the clinical workforce into both the evaluation and identification of opportunities for improvement. For this we interviewed local clinical and management leaders to determine subjective reactions to the changes made. Baseline measures were not available because this process did not start until quarter 4 of 2020 during implementation.

Design

The MC CRP team connected with the SSMC quality team to review their current processes and further define their needs. A project plan was developed, outlining a 12-week engagement that would consist of education, training, demonstrations and coaching to introduce processes, roles and taxonomy of indicators under the care review process (table 1).

The engagement started with weekly meetings to identify SSMC subject matter experts who would become trained in the MC CRP. Staff from SSMC consisted of leaders from the Quality and Patient Safety Committee, Risk and Safety team and the M&M Committee. Mayo Clinic representation included leaders from the Enterprise Mortality Review Subcommittee and the Mayo Clinic Global Quality Consulting Team. Ad hoc staff were brought in for clinical expertise during the SSMC case reviews.

The project’s design consisted of pre-engagement senior leadership meetings to gain agreement about any SSMC committee structural changes required to implement the review process and to ensure agreement in project scope and timeline. These meetings were followed by weekly video engagements between our training team and the SSMC team. An agreed-upon principal step during the pre-engagement meetings was that, in addition to didactic training, working through some case examples would be mutually beneficial. We obtained legal and ethical opinions regarding the ability to access medical information, ensuring that compliance to individual privacies and rights were met.

In this collaboration, SSMC conducted multiple key interventions at its site:

1. Adoption of a care review process for deaths that included perspectives from a nurse and a clinician. This was an evolution of SSMC’s existing review process to provide a more holistic review of the patient’s care.

2. Implementation of a weekly care review huddle, with restructuring of their existing subcommittee to be multidisciplinary, which was then representative of the clinical practice and other non-clinical supportive roles.

3. Incorporation of the use of indicators for improvement opportunities for consistent feedback when closing the loop with the practice.

4. Identification of and feedback on potential opportunities for performance improvement through identified clinician quality improvement champions in each department of the practice.

5. Creation of an education/communication plan for key stakeholders and reviewer roles using change management strategies.

Strategy

N/A.

Lessons and limitations

SSMC’s virtual implementation of a care review process with Mayo Clinic was successful in that it improved the institution’s quality and safety processes. SSMC met its KPI targets and, perhaps more importantly, built a sustainable process allowing for timely feedback to the practice for continuous improvement. Although Mayo Clinic has led several collaborative improvement projects, this was our first international collaboration.

We believe that one key factor for the success of this project was SSMC’s readiness to engage. The institution had clear support from senior leadership and had set this as a priority in the quality strategic plan for the year. This support and readiness may not be evident in all organisations. In a virtual improvement collaborative conducted among 45 Canadian long-term care facilities which aimed to improve appropriateness in antibiotic prescribing,12 an important lesson was the need to evaluate or strengthen readiness for implementation in participating organisations before beginning the collaboratives. In another study of a virtual learning collaborative designed to help establish primary care homes in the Veterans Administration system, the lack of dedicated time for participants by sponsoring organisations led to decreasing satisfaction and outcomes.13 In our case, SSMC requested assistance and was organisationally committed and ready to learn and implement.

SSMC was able to adapt the MC CRP into its own version rapidly because of the organisation’s use of real-time process improvement (eg, use of rapid-cycle tests of change) and change management strategies. Virtual training with established, coordinated, weekly sessions was paramount to completing the training in the weeks defined. It was beneficial to recapitulate the prior week’s training at the start of each new session, helping to update any staff who had questions on the process from the previous week.

Of note, organisations that seek to replicate this type of work should consider allowing time to conduct a thorough current-state assessment. We found that SSMC had a robust M&M process that provided a foundation that could be enhanced to incorporate key aspects of the MC CRP. Learning about the existing process and incorporating it into the new process enhanced staff buy-in and agility to change. Baseline measurements of and problems with SSMC’s existing process provided awareness of gaps to be closed and a way to measure success. A previous virtual learning collaborative in the USA focused on improvement in palliative care-related gaps at the participating local institutions.14 The authors identified that a credible baseline assessment of the individuals and participating organisations, including some detailed knowledge of baseline function and capabilities, would be important for best virtual collaboratives. Other key lessons they cited were identifying the institutional champions for the project, keeping a collaborative cadence, sustaining accountability for progress between meetings and using effective educational methods when transmitting new knowledge.14

Our project benefitted from strong institutional leaders as champions. In addition, we found didactic education in combination with case-based and coached learning to be best, particularly given the potential language barriers. One challenge we encountered was legal issues related to sharing medical information across international borders. It was fortuitous that Mayo Clinic had a newly formed Global Quality Consulting Team that could address legal questions related to document sharing and international EHR access before the current-state assessment. Overall, this was a mutually beneficial learning endeavour for both teams, leading to an effective, ongoing working relationship.

Minimal limitations affected the work with SSMC. The engagement was initially planned to be in person with on-site training, but COVID-19 travel restrictions necessitated a virtual quality improvement project. Although on-site training would have given more insight into culture, operations and flexibility to tailor teaching, the virtual environment still had some advantages. It gave face-to-face familiarity, allowing for a purposeful connection to relay key aspects of SSMC’s organisation as a whole. The training with SSMC was conversation based, which allowed for bidirectional learning in a question-and-answer format. The open format facilitated the identification of topics that needed further exploration and gaps in our training.

Time zone differences and variations in the defined work week limited potential meeting times. It resulted in meetings being scheduled during off-hours for both SSMC and Mayo Clinic staff. Nevertheless, a mutually agreed-upon day and time allowed for greater availability for participation. Lack of coinciding work hours contributed to technical/legal delays in EHR access, which affected case review training. Our sites also used two different EHRs, which required additional work to obtain privileges and chart access; this factor must be considered with any implementation. In addition, SSMC did not use an external database to facilitate their newly developed CRP, but this did not hinder their improvement efforts or ability to successfully identify opportunities.

We also realise that the care review process may be challenging to institute in other areas of the world because of legal, regulatory and cultural differences. However, we propose that some adaptation of this process could be used for greater transparency and advancing safety in many health systems. Cultural readiness is another limitation, but we encourage consideration of implementing a care process review framework to enhance system learning.