The Institute of Medicine’s (1) 1999 report To Err Is Human outlined the challenges of limiting medical error and improving patient care. Examining safety risks, reducing error, and developing initiatives to address problems have become an increasingly critical focus for hospital patient safety and quality programs. Their importance has generated national efforts by many organizations, notably the Institute for Health Care Improvement and the Joint Commission. In many fields of medicine, morbidity and mortality rounds (M&Ms) are an essential component of these programs (2). These rounds are designed to provide timely, structured peer review, prompt reporting, analysis of adverse events, and education in the latest evidence-based practice (2). For medical educators, these rounds cover critical elements of the Accreditation Council of Graduate Medical Education’s (ACGME) core training competencies that must be integrated into every residency program (3).
With roots in surgery and anesthesia dating back through most of the 20th century, M&Ms have been described as the “golden hour of surgical education” (4). They are required for all surgical and anesthesia training programs (3). Although the ACGME does not mandate them for nonsurgical disciplines, they are an integral part of many training programs, as indicated by a recent national survey showing that 90% of internal medicine residencies have M&Ms (2). The integration of the ACGME’s core competencies of interpersonal and communication skills, practice-based learning and improvement, and professionalism into the educational process has been described in an internal medicine M&M setting (5).
Given their importance in patient safety and clinician education, the limitations of M&Ms have been scrutinized (4, 6, 7). The goals of critical self-appraisal, high-spirited discussion, and practice transformation are rarely more than partially achieved (4, 8). These shortcomings have been attributed to problems with the rounds’ format, patient selection, meeting process, and translation of findings into practice (4, 8). Format concerns have included nonstandardized presentations, lack of mandated attendance, or inadequate discussion time (6). Patients without “real” morbidity/mortality have been selected for M&M presentation, which then limits critical self-appraisal (6). Meeting process concerns have centered on stifling of discussion, degeneration into a lecture, and the lack of genuine openness (2). Finally, it has even been suggested that all great ideas are “dismissed at the door” on the way out of M&Ms (4).
Despite these limitations, M&Ms would appear to be a potentially productive venue for self-appraisal and case review that would aid psychiatry departments in their quality improvements and education efforts (7, 9, 10). As with other medical specialties, psychiatry is facing the same challenges from patients, families, hospitals, regulatory agencies, and insurers to enhance patient safety and quality. This makes it somewhat surprising that M&Ms in psychiatry are not required by either the ACGME or hospitals. PubMed, MEDLINE, and PsycINFO searches for psychiatry, morbidity and mortality rounds, sentinel events monitoring, quality assurance, and peer review in articles from 1967–2008 yielded a single earlier report (11) of a general psychiatry M&M and other reports (12–19) that overlap with some elements of this article, but differ appreciably in format, detail, scope, and direction.
This report is the first in the literature to describe the implementation and impact of M&Ms in a child and adolescent psychiatry program.
M&Ms are a component of the safety and quality program at Children’s Hospital in Boston. All hospital departments, including psychiatry, were mandated in 2004 to have monthly M&Ms designed to provide timely review, prompt reporting, and adverse event analysis. These have been guided and are integrated, as are all the departmental M&Ms, by the hospital-wide quality assurance effort from the Program for Patient Safety and Quality.
The psychiatry department encompasses mental health clinicians from psychiatry, psychology, nursing, and social work as well as trainees in psychiatry, psychology, and social work, providing approximately 14,000 outpatient visits, 800 medical consultations, 400 inpatient psychiatry admissions, and 3,000 community-based visits each year.
Psychiatry M&Ms are a monthly, 1-hour conference mandated for all clinicians and trainees. Hospital clinicians outside of psychiatry are invited when they have been directly involved with a patient’s care. To promote interdisciplinary collaboration and review, two senior clinicians from different disciplines (psychiatry and social work) were selected as cochairs.
A standard presentation format was developed focusing on a single case at each round and consisting of a 20-minute uninterrupted presentation, a 25-minute group discussion, and a 10-minute risk-management discussion. At the end, cases are rated by attendees’ consensus on the hospital’s four-level risk-management scale: expected; unexpected, with no identifiable opportunity to improve care; unexpected and potentially avoidable, with a possible opportunity to improve care; and unexpected and definitely avoidable, with an opportunity to improve care.
The cochairs prepared standardized summary reports with the number of attendees, patient demographics, and a description of the event, including clinical findings, diagnoses and interventions, patient outcome, contributing system problems, and recommended actions. The cochairs were also charged to identify recurrent trends in the clinical care and/or hospital systems that would contribute to adverse patient care. These reports were reviewed monthly with the department chief.
The criteria for case selection were mortality (e.g., death due to any cause), substantial morbidity (e.g., suicide attempts), serious complications (e.g., injury from episode of aggression), major missed diagnoses (e.g., agitation in a depressive episode), treatment errors (e.g., wrong medication prescribed), and challenging/problematic cases that resulted in a substantial change in a patient’s course and/or system problems. These criteria are routinely monitored by clinical service directors and the department chief under the guidelines of the hospital’s Program for Patient Safety and Quality. Using these guidelines, the cochairs selected the most problematic cases that came to the department chief and/or the clinical service directors. Less problematic cases continued to be reviewed in the primary settings by clinical service directors. In months in which a major case was not readily identified, a challenging case was selected from one of the four department clinical services on a rotating basis.
The rounds were structured to be constructive and collegial. This meant that no individual clinician was to be subjected to derogatory or demeaning treatment. The M&Ms were conducted as a peer-reviewed activity, subject to the statutes protecting peer-reviewed activities that promote fully protected disclosure.
Understanding that nonphysician clinicians would be unfamiliar with the M&M format and that even most psychiatrists would have had few M&M experiences, the cochairs emphasized the elements of M&Ms that are different from didactic conferences or process-oriented case discussions. Initially, many participants found that the familiar “processing” of cases was sacred ground that they were reluctant to give up. The cochairs repeatedly acknowledged the importance of processing case-generated feelings, but that these were not the focus for M&Ms unless they could be tied directly to the medical error. Instead, the cochairs reemphasized that M&Ms were participatory discussions to openly and critically examine clinical thinking and action in the context of challenging cases.
The rounds start with a concise summary of what happened and then focus on what was behind the interventions or actions and what might have been done differently. The why behind every inquiry or suggestion is stressed, and participants are asked to explicitly state the underlying reasoning behind their questions and suggestions. The goal of each meeting is to examine practice while looking for opportunities to improve care. These principles are restated and confidentiality is reviewed at the start of each M&M.
Psychiatry M&Ms Impact Assessment
To understand the impact of psychiatry M&Ms on individual clinicians, in January 2006 and 2007 all participants were asked to anonymously complete an evaluation survey that elicited quantitative data rated on a 5-point Likert scale ranging from 1 (“disagree strongly” or “poor”) to 5 (“agree strongly” or “excellent”). The survey included questions regarding the M&M’s overall usefulness, constructiveness/collegiality, meeting of expectations, impact on clinical practice, quality of case discussions, and ratings of format. Mean ratings between the 2 years were compared using independent t tests. To assess impact on clinical practice, the cochairs recorded changes in patient care that occurred as a direct result of M&Ms, as reported by clinical service directors.
Between July 2005 and May 2007, 19 psychiatry M&Ms were held, with the attendance ranging from 50 to 80 clinicians at each. The mean patient age was 15.2 years, and 70% of the patients were female. The cases came from inpatient (35%), emergency room (35%), outpatient (15%), medical consultation (15%), and community (10%) sites. The primary presenting problems were self-injury/suicide attempts (35%), aggression/assaults (35%), other care failure (25%), and diagnostic concerns (5%). Deaths occurred in three cases, one in a motor vehicle crash of an active outpatient and two recent patients who were under care at other facilities at the time of their demise.
On the hospital’s four-level risk management rating, 80% of the cases were rated as being unexpected and potentially or definitely avoidable with an opportunity for improvement, 10% as unexpected with no opportunity for improvement, and 10% as having expected problems.
As part of the continuing review for trends or patterns that adversely impact patient care, four recurrent problem areas were identified: diagnostic and formulation errors, communication problems, systems-based problems, and class and cultural misunderstandings (Table 1).
With a 75% survey completion rate each year, the composition of respondents in 2006 (n=63) was 44% physicians, 32% psychologists, and 24% social workers. Of the respondents in 2007 (n=57), 41% were physicians, 33% were psychologists, and 26% were social workers. Of trainees, 35% responded in 2006 and 38% in 2007.
The overall usefulness, discussion quality, and rounds format were rated highly, with the mean values both years averaging around 4.0 on the 5-point Likert scale (Table 2). Discussion quality improved significantly in the second year. Impact on expectations and clinical practice were rated lower. The latter was noted despite changes in department practices. Constructiveness/collegiality was significantly lower in the second year, with the decline appearing to be related solely to social workers having a mean ratings decline from 3.8 to 2.8 (p<0.01). Examination of constructiveness/collegiality and each of the other questions by discipline, by trainees compared with staff, and by year yielded no other additional significant differences.
Morbidity and mortality rounds (M&Ms) appear to be an innovative venue for self-appraisal and case review that can aid psychiatry programs in patient safety and education efforts. Possible improvement opportunities were found in more than 80% of the cases presented. The four problem areas (diagnostic and formulation errors, communication problems, system-based problems, and class and culture misunderstandings) were not mutually exclusive; problems occurred in different combinations with varying emphasis. Identifying these areas led the department to take corrective actions that positively and significantly impacted patient care. Because the M&Ms were mandated for all professional staff and supported and attended by the department chair and the heads of nursing, social work, and psychology, these changes were readily implemented and then monitored by the respective division chief.
Identifying and addressing recurrent problem areas are congruent with the ACGME core competencies of interpersonal and communication skills, practice-based learning and improvement, and professionalism (3). As a result, these core skills were invariably at center stage during some point of every M&M. The rounds also provided senior staff a unique opportunity to model these critical skills for trainees and junior staff. In particular, the acknowledgment of errors and/or differences in judgment by senior staff can be powerful in modeling analytic thinking and transparency in caring for patients.
The evaluation suggested that multidisciplinary psychiatry M&Ms are useful and educational. The format and accompanying discussion were perceived as being of high quality. The overall usefulness to clinicians themselves was found to be high. Perceived collegiality declined in the second year. This change may have been catalyzed by a shift in M&M facilitation, in which the cochairs were more insistent that each participant explain the why behind questions or suggestions. Although this led to more rigorous discussion (e.g., significant improvement in discussion quality), the shift may have left some participants feeling more “on the spot” and less “safe.”
The format of critical case review generally is quite foreign to most mental health clinicians. Breaking the responses down by discipline, a shift in the social workers’ perceptions appeared to underlie the reduction in overall perceived collegiality. This was reviewed by the senior psychiatrist and social work cochairs without clear etiologic findings. Although having senior clinician cochairs appears to be a useful innovation, leaders of these rounds must just be cognizant of the differences in training and clinical models as they facilitate case discussions.
Despite substantial changes in departmental operations, the rounds only had a moderate impact on individual clinical practice. There also were only moderate ratings on how well M&Ms met individual expectations. Together, these findings support the premise that M&Ms are challenging for clinicians perhaps because of their reticence to formally outline individual diagnostic rationales and treatment approaches in a group setting. The challenge could be a product of local hospital or general health care system forces that dissuade personal reflection and public culpability.
The implementation of M&Ms within a children’s hospital setting may limit generalization to other psychiatry settings. The problem area determination may be limited given that the analysis was impressionistic and qualitatively based on monthly reviews between the cochairs and the department chief. Of course, these problem areas are meant to be guidelines rather than absolutes. Although the evaluation ratings were generally strong, they may have been limited by being structured to elicit only initial feedback, feasibility, safety, and impact rather than comparative value and longer-term impact. In this manner, they may have introduced some positive bias in the evaluation’s format, but we believe that they still demonstrate M&Ms’ initial feasibility, safety, and value. Clearly, future surveys will be needed to study M&Ms’ long-term impact and comparative value to other modalities as an educational and quality assurance activity. Although the survey had an excellent completion rate, result differences might have been reflective of attendee changes between years. Although the case finding mechanisms appeared to be successful in identifying every case with a significant adverse outcome, there may have been a selection bias that missed cases. Further study is needed to document more comprehensively that the ACGME competencies are integral elements of rounds and the long-term impact on staff and trainee clinical practices.
We believe that M&Ms have significant potential to be successfully and safely utilized by psychiatry programs in helping to achieve their risk management, clinical care, and educational goals. We have outlined an adapted psychiatric format to accomplish this in a pediatric psychiatry setting, which we believe may be directly extended to general psychiatry settings. The identification of the specific problem areas of diagnosis and formulation errors, communication problems, system-based problems, and culture misunderstandings can serve as scaffolding to examine problem cases, improve patient care, address ACGME core competencies, and enhance medical education in psychiatric departments.
At the time of submission, the authors declared no competing interests.