Implementation status of morbidity and mortality conferences in Swiss hospitals: a national cross-sectional survey study.

OBJECTIVE: To determine the implementation status and current practice of morbidity and mortality conferences (M&MCs) in Switzerland.
DESIGN: A national cross-sectional online survey was conducted in spring 2017. The questionnaire focused on overall goals, structure and procedures of hospital M&MCs. Further topics included satisfaction, perceived effectiveness and support requirements.
SETTING: A total of 913 chief physicians of surgery and internal medicine, and specialist fields of obstetrics and gynaecology, anaesthesiology and intensive care from Swiss acute care hospitals were invited to the survey. 321 completed the questionnaire, resulting in a 35.2% response rate. PARTICIPANTS: Chief or senior physicians in charge of the M&MCs in their department. INTERVENTION: No intervention. MAIN OUTCOME MEASURES: Numbers and percentages of M&MCs within the surveyed disciplines fulfilling certain characteristics and procedural features.
RESULTS: Among 321 respondents, the majority are conducting M&MCs in their departments. Within and between the medical disciplines considerable heterogeneity was found in structural and procedural features of M&MCs. Only a small part of the reported M&MCs is following a systematic approach and meeting recommended procedural features. Although the respondents are satisfied and perceive the M&MCs as an efficient tool, they agree that there is a need for professionalization and standardization.
CONCLUSION: M&MCs are widely used to promote medical education, patient safety and quality improvements. However, the term M&MC seems to cover different types of meetings. Although the overall goals are similar, various types of M&MCs are used in practice and different objectives are pursued. Tools such as checklists, guidelines and templates are considered helpful.

Introduction

Morbidity and mortality conferences (M&MCs) have a long tradition as a forum for continuing medical education in English speaking countries. Traditionally, they were used by surgeons to retrospectively discuss medical errors, complications or unexpected deaths [1]. In recent decades, M&MCs have been recognized internationally as a powerful tool for promoting patient safety [2-4]. Other medical disciplines such as obstetrics and gynaecology, intensive care [5], and internal medicine [6] have adopted M&MCs. Additionally, M&MCs are becoming more and more interdisciplinary and interprofessional, connecting physicians from different specialties, nurses and other staff to focus on common learning from medical errors and complications [7].

International studies identified ambiguities in goals and tasks of M&MCs, heterogeneity in their implementation and structure, an absence of standardized mechanisms to identify and address errors, as well as a lack of adequate and timely follow-up [1, 6, 8–11]. Differences were found across departments [8], for instance between internal medicine and surgery [1], as well as within specific disciplines [14, 15]: differences were detected with respect to the number and type of cases discussed during M&MCs [12, 13] as well as in the way cases were presented [1] and reviewed [8]. Aboumatar et al. found significant variation in the organization of M&MCs even within a single institution [8]. Both content and goals of conferences varied across departments. Among 12 departments, there were several approaches in reviewing cases. Most conferences focused on issues of medical management, but less than half concentrated on patient safety or quality issues. Pierluissi et al. [1] found differences between internal medicine and surgery conferences with respect to the format, numbers and frequency of adverse events and errors in cases presented, time spent on the case presentation and discussion.

While M&MCs originally focused on improving individual practice by critically reflecting on completed cases, they are now increasingly taking a system-oriented perspective. Standardized and structured approaches to case discussion as well as error and incident analysis (e.g. Fishbone analysis) [3] are being integrated to detect systemic problems, e.g. in the organization of care or the interaction of medical staff and physicians. Other studies showed that standardized presentations can have an impact on participant satisfaction [16] and M&MCs using a formalized framework were perceived as being more effective [9, 13].

In Switzerland, M&MCs are part of the continuing education programs of most medical disciplines recognized by the Swiss Institute for Continuing Medical Education [17]. However, little is known about their implementation status, the characteristics of M&MCs within different medical disciplines and the pursued goals. The aim of this study was to determine the implementation status and spread of M&MCs in Switzerland. We focused on the organizational structure, procedures and characteristics of M&MCs in Swiss acute care hospitals. Chief physicians served as our key informants for obtaining data about M&MCs at their departments.

Methods

A cross-sectional survey was conducted in spring 2017.

Survey instrument

Since this was the first national survey related to M&MCs, a new questionnaire was designed. The survey was based on an extensive review of the available literature, assessments of national experts and internationally established guidelines, e.g. the Recommendations of the German Medical Association [18], the Ottawa M&M Model [19], or Models from Surgery in England [20, 21]. The questionnaire comprised three sections: The participants were first asked whether a M&MC is currently implemented at the clinic or department. Participants without M&MCs were asked whether they would be interested in introducing M&MCs. Participants with existing M&MC were administered the remaining survey sections. The second section included questions about the overall goals as well as structure and process characteristics of M&MCs (29 items). In the third section, the satisfaction and perceived effectiveness of the M&MCs were assessed. The questionnaire ended asking whether and how M&MC could be improved and the sentence ‘M&MCs as a tool to learn from errors could bring more benefit if …’ which the respondents had to complete. At the end of the survey, socio-demographic questions were asked. Answer options were binary or categorical and some questions provided free-text fields. The final questionnaire had 44 items in total. The German language draft version of the survey was tested for functionality, comprehension and acceptance in a small group of senior physicians. The resulting feedback was integrated in the survey. As Switzerland is a multilingual country the survey developed in German was subsequently translated into French and Italian by professional translators. The translations were approved by native speakers. The survey was programmed as a secured online survey with individual access codes.

Sample and procedures

The target group included chief physicians of surgery, internal medicine, obstetrics and gynaecology, anaesthesiology and intensive care of all Swiss acute care hospitals. An invitation letter was sent by post and e-mail to a total of 913 chief physicians. The letter introduced the study and its aims and provided them with the survey URL and personal access code. Chief physicians were allowed to delegate survey participation to the person in charge of the M&MCs in their department. An electronic reminder was sent two weeks after the initial invitation. The survey was officially sent, fielded and coordinated by the Swiss Patient Safety Foundation and co-signed by the Swiss Medical Association (FMH). The study was exempt from review by the Cantonal Ethics Board (BASEC-Req-2017-00 325). Survey participation was considered informed consent.

Data analysis

Results are presented as numbers and percentages (%). To test for differences in characteristics of reported M&MCs between medical disciplines Chi2 test and Fishers exact test (for cell counts ≤5) were used as appropriate. The Kruskal–Wallis test was used to test for differences in ordinal outcomes between medical disciplines. A P-value <0.05 was considered statistically significant. Responses to open-ended questions were analysed for content and qualitatively coded by two of the authors. Coding was discussed within the research team.

Results

Sample

Of the 913 invited physicians, 321 completed the survey, resulting in a 35.2% response rate. Compared to the entire sample, surgical disciplines (35.8% vs. 32.2%) and anaesthesiology/intensive care (25.2% vs. 21.5%) disciplines are slightly overrepresented whereas internal medicine disciplines (32.7% vs. 37.4%) and obstetrics/gynaecology (6.2% vs. 9.0%) are underrepresented among survey participants (Chi2 = 0.006). The majority of respondents are chief physicians (77.5%) and senior physicians (21.3%). Sample characteristics are presented in Table 1.

Table 1

Demographics of survey participants

		Total sample N = 321 (100%)	M&MCs currently implemented N = 223 (69.5%)	M&MCs currently NOT implemented N = 98 (30.5%)
		n	n (%)	n (%)	P-value
Survey language	German	232	168 (72.4)	64 (27.6)	0.040a
	French	73	48 (65.8)	25 (34.2)
	Italian	16	7 (43.8)	9 (56.3)
Function+	Chief physician	248	170 (68.5)	78 (31.5)	0.348b
	Senior physician	68	51 (75.0)	17 (25.0)
	Other physician	4	2 (50.0)	2 (50.0)
Gender+	Female	37	26 (70.3)	11 (29.7)	0.935a
	Male	283	197 (69.6)	86 (30.4)
Hospital categories+	University hospital	67	55 (82.1)	12 (17.9)	<0.001a
	General hospital with ≥500 beds	48	38 (79.2)	10 (20.8)
	General hospital with 125–499 beds	117	85 (72.7)	32 (27.3)
	General hospital with ≤124 beds	63	39 (61.9)	24 (38.1)
	Special clinics	24	6 (25.0)	18 (75.0)
Legal structure+	Public hospital	272	198 (72.8)	74 (27.2)	0.009a
	Private hospital	47	25 (53.2)	22 (46.8)
Medical disciplines	Surgery	115	97 (84.4)	18 (15.6)	<0.001b
	Internal medicine	105	55 (52.4)	50 (47.6)
	Anaesthesiology and intensive care	81	56 (69.1)	25 (30.9)
	Obstetrics/gynaecology	20	15 (75.0)	5 (25.0)
Number of beds in department+	Up to 20	52	32 (61.5)	20 (38.5)	0.159b
	21–40	64	46 (71.9)	18 (28.1)
	41–60	50	38 (76.0)	12 (24.0)
	61–80	33	28 (84.8)	5 (15.2)
	Over 80	53	37 (69.8)	16 (30.2)
	Not applicable	67	42 (62.7)	25 (37.3)

+Not adding up to n = 321 in the total sample due to missing values.

aChi2 test for differences between respondents with/without implemented M&MC.

bFisher exact test for differences between respondents with/without implemented M&MC when cell counts are ≤5.

The majority of participating physicians indicated that M&MCs are implemented in their clinic or department (69.5%). From the 98 participants, who do not conduct M&MCs, the majority is interested (44.3%) or rather interested (43.3%) in incorporating M&MC into their future practice. There are several differences between respondents with and without M&MCs (Table 1).

Description of M&MCs in Swiss hospitals

The following analysis is based on the 223 individuals, who confirmed that M&MCs are conducted in their clinic or department. Table 2 summarizes the main characteristics of M&MCs as currently implemented.

Table 2

Characteristics by medical disciplines

	Total N = 223		Internal medicine N = 55		Anaesthesiology & intensive care N = 56		Surgery N = 97		Obstetrics/gynaecology N = 15
	n	(%)	n	(%)	n	(%)	n	(%)	n	%	P-value
Frequency+
Weekly/fortnightly++	35	(15.8)	4	(7.3)	4	(7.1)	26	(27.1)	1	(6.7)	0.001a
Monthly	46	(20.7)	7	(12.7)	11	(19.6)	24	(25.0)	4	(26.7)
Quarterly	98	(44.1)	25	(45.5)	28	(50.0)	36	(37.5)	9	(60.0)
Rarely than quarterly/irregular++	43	(19.4)	19	(34.6)	13	(23.2)	10	(10.4)	1	(6.7)
Duration
≤30 min++	50	(22.4)	9	(16.4)	6	(10.7)	33	(34.0)	2	(13.3)	0.0292a
31–45 min	58	(26.0)	17	(30.9)	15	(26.8)	22	(22.7)	4	(26.7)
46–60 min	84	(37.7)	20	(36.4)	26	(46.4)	32	(33.0)	6	(40.0)
>60 min	31	(13.9)	9	(16.4)	9	(16.1)	10	(10.3)	3	(20.0)
Number of attendees+
3–10 attendees	73	(32.7)	12	(21.8)	23	(41.0)	36	(37.1)	2	(13.3)	0.0270a
11–20 attendees	83	(37.2)	20	(36.4)	19	(33.9)	37	(38.1)	7	(46.7)
21–30 attendees	36	(16.1)	11	(20.0)	6	(10.7)	14	(14.4)	5	(33.3)
>30 attendees	29	(13.0)	12	(21.8)	7	(12.5)	9	(9.3)	1	(6.7)
Estimated participation rate of invited staff+
1–25%	15	(7.0)	7	(12.7)	7	(14.0)	0	(0.0)	1	(7.7)	< 0.001a
26–50%	60	(28.0)	13	(23.6)	24	(48.0)	18	(18.8)	5	(38.5)
51–75%	76	(35.5)	21	(38.2)	15	(30.0)	36	(37.5)	4	(30.8)
76–100%	63	(29.4)	14	(25.5)	4	(8.0)	42	(43.8)	3	(23.1)
No. of cases per M&MC
1 case	59	(26.5)	22	(40.0)	15	(26.8)	19	(19.6)	3	(20.0)	0.225a
2 cases	75	(33.6)	11	(20.0)	21	(37.5)	39	(40.2)	4	(26.7)
3 cases	53	(23.8)	13	(23.6)	18	(32.1)	18	(18.6)	4	(26.7)
≥4 cases++	36	(16.1)	9	(16.4)	2	(3.6)	21	(21.6)	4	(26.7)
Time per case+++
≤19 min++	70	(31.4)	12	(21.8)	11	(19.6)	42	(43.3)	5	(33.3)	0.0012a
20–29 min	49	(22.0)	13	(23.6)	10	(17.9)	22	(22.7)	4	(26.7)
30–44 min	61	(27.4)	17	(30.9)	19	(33.9)	23	(23.7)	2	(13.3)
≥45 min	43	(19.3)	13	(23.6)	16	(28.6)	10	(10.3)	4	(26.7)
Time per case for open discussion
≤10 min++	76	(34.1)	13	(23.6)	17	(30.4)	40	(41.2)	6	(40.0)	0.466a
11–15 min	70	(31.4)	20	(36.4)	21	(37.5)	25	25.8	4	26.7
16–20 min	46	(20.6)	15	(27.3)	9	(16.1)	20	20.6	2	13.3
≥20 min	31	(13.9)	7	(12.7)	9	(16.1)	12	12.4	3	20.0
Presenter
Chief physician	23	(10.3)	8	(14.6)	7	(12.5)	7	(7.2)	1	(6.7)	0.120b
Senior physician	98	(44.0)	24	(43.6)	29	(51.8)	38	(39.2)	7	(46.7)
Resident	88	(39.5)	19	(34.6)	15	(26.8)	49	(50.5)	5	(33.3)
Other	14	(6.3)	4	(7.3)	5	(8.9)	3	(3.1)	2	(13.3)
Role allocation
(A) 1 person chairs, moderates and presents (all in one)	23	(10.3)	9	(16.4)	4	(7.1)	9	(9.3)	1	(6.7)	0.309b
(B) 1 person chairs and moderates & 1 or more others persons present	153	(68.6)	39	(70.9)	39	(69.6)	65	(67.0)	10	(66.7)
(C) 1 person chairs & 1 person moderates & 1 or several persons present	27	(12.1)	3	(5.5)	9	(16.1)	11	(11.3)	4	(26.7)
(D) 1 person chairs & 1 or several other persons present (no moderation)	20	(9.0)	4	(7.3)	4	(7.1)	12	(12.4)	0	(0.0)
Types of improvement measures
Individual	32	(14.4)	7	(12.7)	7	(12.5)	16	(16.5)	2	13.3	0.965b
Local	133	(59.6)	33	(60.0)	36	(64.3)	56	(57.7)	8	53.3
Systemic	58	(26.0)	15	(27.3)	13	(23.2)	25	(25.8)	5	33.3

+May not sum up to total N due to missing value. ++Merged over two categories with few responses. +++Post-hoc developed categories based on free-text responses.

aKruskal–Wallis test for differences in ordinal outcomes.

bFisher exact test for differences between respondents from different medical disciplines when cell counts are ≤5.

M&MCs goals

Respondents could choose three out of five M&MC goals representing individual and organizational learning. The most commonly reported overall goals of M&MCs are ‘Preventing recurrence of errors’ (95.5%), ‘Identifying problems in the processes’ (83.0%) and ‘Improving collaboration between professionals and departments’ (60.5%) (Figure 1). These goals all relate to organizational learning. Goals representing individual learning, such as ‘Expanding individual knowledge’ (39.5%) and ‘Learning about rare diseases’ (6.3%) were reported less frequently. 57.9% of respondents reported a combination of both individual and organizational learning goals.

Figure 1

Overall goals.

There are considerable differences in reported goals between disciplines (Figure 1). ‘Preventing recurrence of errors’ is essential for representatives of surgery (100%) but less so for representatives of other medical disciplines (92.1%, P = 0.005). ‘Improving collaboration’ was most widely chosen by responders working in anaesthesiology and intensive care (75.0%) compared to surgery (49.5%), internal medicine (63.6%) or obstetrics/gynaecology (66.7%).

M&MCs characteristics

Main characteristics of current M&MCs are listed in Table 2. Most departments hold conferences monthly or quarterly and commonly discuss two to three cases. Overall, chief and senior physicians (92.8% and 97.3% respectively) and residents (94.6%) belong to the most frequent M&MC attendees. Almost half of the departments include participants from other disciplines for their M&MCs (e.g. pharmacists, pathologists or radiologists). In 46.2% of the departments, nurses usually attend M&MCs.

The respondents could choose up to three types of cases that are usually presented in M&MCs. Typically presented cases in M&MCs are complications (78.0%), unexpected mortality (50.7%) and severe illnesses or progressions (28.7%). Some departments include incidents and critical events without or with patient harm (27.8% vs. 18.4%). Cases are selected based on problems in cooperation (39.5%), deficits in organizing care (15.7%) and gaps in medical knowledge or deficiencies in clinical skills (24.2%). Rare diseases (6.3%) and cases with little potential for conflict between the involved parties (0.9%) are only rarely selected for presentation.

Nearly half of respondents spend more than 30 min per case (46.6%). The amount of time dedicated to open discussion per case is up to 10 min in 34.1% of the departments. Only in 13.9% of the departments the discussion is lasting over 20 min. The most common role allocation reported includes one person who chairs and moderates, and one or more persons who present (68.6%). Less frequently a person chairs the M&MC, another person moderates, and one or more other persons present the cases: Only 12.1% reported designated roles for moderator and chair. Nearly all sites assign a moderator to the session (91.0%), but this function is commonly combined with chairing the M&MC (78.9%). Most derived improvements include local measures that optimize a (sub-) process in one area and locally (59.6%). Systemic measures that optimize a (partial) process for the entire hospital and have a global or systemic effect are less frequent (26.0%). Individual measures aimed at changing the behaviour of individual employees are only rarely derived (14.4%).

Procedural features of M&MCs

Procedural features of existing M&MCs are reported in Table 3. Some departments have clear criteria on how to select cases (41.3%). Significant differences between clinical disciplines were observed in case selection and case preparation. Out of 203 respondents only 59 reported that moderators are trained (29.1%). While almost all respondents stated that measures for improvement are defined during M&MCs (91.5%), across disciplines only a minority used any models, guidelines and key questions for case analysis and discussion. Some respondents analyse and present cases according to a theoretical model or guidelines. The SBAR-Communication-Model [22], the guidelines of IQM [23] and the London Protocol [24] were named several times. Twenty-three respondents mentioned own existing guidelines.

Table 3

Procedural features of M&MCs

For the M&MC in our department	Total		Internal medicine		Anaesthesiology & intensive care		Surgery		Obstetrics/gynaecology
	N = 223		N = 55		N = 56		N = 97		N = 15
	n	(%)	n	(%)	n	(%)	n	(%)	n	%	P-value
Cases are selected by defined criteria	92	(41.3)	17	(30.9)	12	(21.4)	52	(53.6)	11	(73.3)	<0.001b
Cases are prepared and reviewed in a standardized way	126	(56.5)	25	(45.5)	23	(41.1)	65	(67.0)	13	(86.7)	<0.001b
Cases are analysed and presented using a model or guidelines	54	(24.2)	12	(21.8)	10	(17.9)	27	(27.8)	5	(33.3)	0.417b
Cases are discussed along key questions	64	(28.7)	22	(40.0)	11	(19.6)	26	(26.8)	5	(33.3)	0.108b
Person who moderates received training in moderation	59	(29.1)	10	(19.6)	15	(28.8)	26	(30.6)	8	(53.3)	0.086a
Improvement measures are defined during the M&MC	204	(91.5)	52	(94.6)	50	(89.3)	87	(89.7)	15	(100.0)	0.509b
Results are recorded in a protocol or written summary	82	(36.8)	22	(40.0)	17	(30.4)	34	(35.1)	9	(60.0)	0.184a
Results are disseminated internally	124	(55.6)	30	(54.6)	31	(55.4)	51	(52.6)	12	(80.0)	0.264b
Attendees receive feedback on the implementation of the defined measures	86	(38.6)	21	(38.2)	20	(35.7)	34	(35.1)	11	(73.3)	0.045b
Suggestions for improvement of the M&MC by staff are actively sought	122	(54.7)	29	(52.7)	34	(60.7)	47	(48.5)	12	(80.0)	0.098b

aChi2 test for differences between respondents from different medical disciplines.

bFisher exact test for differences between respondents from different medical disciplines when cell counts are ≤5 (including not shown cells of ‘no responses’).

Satisfaction, perceived effectiveness and improvement potential

The respondents are mostly satisfied (68.6%) or very satisfied (9.9%) with the current M&MCs, while 20.6% are not satisfied and 0.9% very unsatisfied. Overall, respondents rated the conferences as effective (65.5%) or very effective (20.2%) in improving patient safety. 13.9% find them little effective and 0.5% not effective at all. The proportion of participants perceiving the M&MC effective and very effective is higher among those who have certain procedural features, namely defined criteria for case selection (94.6% vs. 79.4%, P = 0.001), standardized preparation (92.1% vs. 77.3%, P = 0.002) and guidelines (96.3% vs. 82.3%, P = 0.01).

The physicians were asked to complete the sentence ‘M&MCs as a tool to learn from errors could bring more benefit if …’. Answers were sorted into categories by consensus of the authors. Answers most frequently related to ‘an open learning and error culture’ (18.1%), followed by ‘better and interdisciplinary participation’ (12.8%) and ‘a consistent implementation of the M&MCs itself and of the defined improving measures’ (12.1%). ‘Higher frequency’ (10.7%) and ‘a systematic approach’ (10.1%) were also proposed as to generate more benefit from M&MCs. At the end of the survey, the physicians were asked if their M&MC could be improved and how this may be achieved. 65.0% of the respondents agreed that they see potential for improvement and suggested various materials as helpful (Table 4).

Table 4

Suggestions for helpful M&MC tools

	n	%
Checklist for preparation	148	66.4
Guidelines for conducting M&MCs	131	58.7
Templates for presentation	102	45.7
Training offers for moderation	88	39.5
Minimal standards	72	32.3
Others	25	11.2

Discussion

Our study is the first to provide data on the current implementation and design of M&MCs in Switzerland. The results show that M&MCs are widely implemented within the surveyed medical disciplines in Swiss hospitals. Considering the relatively high response rate in chief physicians and the fact that nearly 87.6% reported intentions to implement M&MCs indicate a considerable interest in the topic.

Although M&MCs seem to be established in Swiss Hospitals they are variable in their structure. Some clinics follow a systematic pattern while others are determined by personal preferences of the chair or availability of time and resources. As others, we find considerable heterogeneity in structural and procedural features of implemented M&MCs within and between the medical disciplines [1, 8, 13, 16]. The differences in frequency, duration and number of cases discussed confirm that different approaches exist and that M&MC is an umbrella term covering several types of meetings. Chief physicians seem to pursue a variety of goals with the M&MC, focusing on organizational learning goals more than on individual learning goals. Structural and procedural variation could be explained by the fact that M&MCs as a whole are currently shifting from the traditional educational instrument for medical trainees towards improving patient safety [2, 7, 13]. However, such a change in focus needs to be accompanied by adequate structural and procedural support. For example, a M&MC aimed at organizational learning and involving different professional groups could probably not be held weekly with four cases discussed within 30 min. Importantly, a less formalized approach may be useful to begin to establish a culture of open discussion about errors within a profession. The unit’s goals must be reflected in its culture as well as in its structural and procedural features and should be developed accordingly and relative to another.

Many of the M&MCs reported about have characteristics that do not match recommendations in the recent literature. For example, nearly every second department covered in our study holds the M&MCs quarterly, while weekly, fortnightly or monthly settings have been recommended [6, 8, 9, 22]. However, regular meetings may be more important than higher frequency of M&MCs [25]. Standardized procedures are not yet common in Swiss Hospitals. There is room for an increased adoption of clear case selection criteria, standardized models, discussion along key questions, protocols and consistent follow-up. It is noteworthy that nearly 50% of departments do not yet have clear criteria and procedures for case selection. A well-defined selection procedure is important to ensure that relevant cases are not missed, to increase transparency and reliability, and to avoid that M&MCs are misused to expose colleagues. Lack of knowledge and of time, as reported by other studies [15], could be a reason why analytical methods are underused. Many authors recommend using a moderator in M&MC [5, 22], for example, a senior physician or an experienced resident [26]. Based on our data we can identify potential for improvement concerning the use and training of moderators: In our sample, M&MCs were mostly moderated and chaired by the same person in a dual role and training of moderators was rather uncommon. Less than a quarter of respondents use guidelines on the analysis and presentation of cases. It is interesting to note that we observed an association of these procedural features with participants’ perceived effectiveness of the M&MCs, suggesting that chief physicians see the benefits of a structured approach. This result is consistent with the literature [15, 16]. Several studies reported higher satisfaction scores associated with M&MCs incorporating structured formats and case analysis [3, 9, 27, 28]. Of course, the causality of this association and its direction remain unclear.

Overall, the survey shows that the majority of respondents are satisfied with the current implementation of their M&MC and perceive the instrument as effective for patient safety. Still, two-thirds of respondents felt that there was room for improvement for the M&MC in their department. Respondents in our survey consider tools such as a checklist for preparation, guidelines and templates as conducive to increased benefits.

Limitations

The main limitation of this survey study is the self-reported nature of our data. The study has standard limitations associated with a response rate below 100%. Nonetheless, the response rate was relatively high considering the target group. Given that the chief physicians are describing their own M&MC results may also be subject to social desirability bias. Although chief physicians are best positioned to describe the structural and procedural characteristics, a more reliable assessment would come from direct observation of M&MCs or surveying all participants. Our study does not provide information on achievement of specific goals or impact measures. Because of the heterogeneity of M&MC models and the reported implementation, comparisons across the medical disciplines are limited and should be made with caution. Despite these limitations, we identify a positive trend in the development and acceptance of M&MCs.

Conclusion

No standard approach to the M&MC currently exists in Switzerland. The M&MCs varied widely in their goals and characteristics. Different formats are established depending on the medical disciplines and settings. This survey provides an overview of M&MCs in Swiss hospitals, and may be a starting point to discuss and develop measures for improvement and increased effectiveness. We find that some characteristics are associated with a perceived higher effectiveness of M&MCs and recommend structured approaches and interdisciplinary audience to improve M&MCs and development towards a system-wide approach. M&MCs should be implemented regularly and in a more structured manner. The implementation of procedural features can be improved, for example, by discussing cases along key questions, by training moderators or recording the results in a protocol or written summary. Respondents consider checklists, guidelines for conducting M&MCs and templates for presentation as helpful tools. We recommend such tools to institution who like to support their medical professionals in conducting M&MCs. Institutions who would like to assess their own practices could use our list of features to check the extent the procedural features have been implemented in their departments.

Further studies are needed to investigate the perception of all participants of M&MCs, the impact of M&MCs on patient safety and how to improve and standardize the conference.