Standardization of Trauma, General Surgical Morbidity and Mortality Conferences: Development and Dissemination of a “Toolkit” in Peru

Background Morbidity and Mortality (M&M) conferences allow clinicians to review adverse events and identify areas for improvement. There are few reports of structured M&M conferences in low- and middle-income countries and no report of collaborative efforts to standardize them. Methods The present study aims to gather general surgeons representing most of Peru’s urban surgical care and, in collaboration, with trauma quality improvement experts develop a M&M conferences toolkit with the expectation that its diffusion impacts their reported clinical practice. Fourteen general surgeons developed a toolkit as part of a working group under the auspices of the Peruvian General Surgery Society. After three years, we conducted an anonymous written questionnaire to follow-up previous observations of quality improvement practices. Results A four-component toolkit was developed: Toolkit component #1: Conference logistics and case selection; Toolkit component #2: Documenting form; Toolkit component #3: Presentation template; and Toolkit component #4: Code of conduct. The toolkit was disseminated to 10 hospitals in 2016. Its effectiveness was evaluated by comparing the results of surveys on quality improvement practices conducted in 2016, before toolkit dissemination (101 respondents) and 2019 (105 respondents). Lower attendance was reported by surgeons in 2019. However, in 2019, participants more frequently described “improve the system” as the perceived objective of M&M conferences (70.5% vs. 38.6% in 2016; p < 0.001). Conclusion We established a toolkit for the national dissemination of a standardized M&M conference. Three years following the initial assessment in Peru, we found similar practice patterns except for increased reporting of “system improvement” as the goal of M&M conferences.


Introduction
Surgical diseases affect all ages, and injury, an emergency surgical problem, is the leading cause of death for people 45 years or younger and 90% of these occur in low-middleincome countries (LMICs), where 85% of the world's population live [1][2][3][4]. It is estimated that 34-38% of all injury deaths in LMICs are preventable [3], if fatality rates among severely injured patients could resemble those in high-income countries (HICs) [3]. Furthermore, a study in 2003 compared trauma systems in HICs and LMICs and found that most of the staff in LMICs had less formal trauma training [4]. Trauma quality improvement programs (TQIPs) include structured assessment of patient care and patient outcomes to identify system improvements that may reduce preventable deaths and improve the processes of care. System referring to those services responsible that improve, maintain or restore health of individuals as per the World Health Organization (WHO) [5].
The WHO and the International Association for Trauma Surgery and Intensive Care (IATSIC) have developed guidelines that summarize the essential components of trauma care across resource settings, including TQIPs. In a pilot study with 23 level I and II trauma centers, TQIPs were shown to be feasible and associated with a significant improvement in risk-adjusted mortality for blunt singleand multi-system trauma [1,4]. Similarly, a WHO-IATSIC collaborative review showed that most TQIPs led to improvement in patient outcomes [6]. However, most of these data come from HICs [1,6].
Morbidity and Mortality (M&M) conferences, a fundamental component of TQIPs, are opportunities for providers to review deaths and complications with a focus on system-wide improvement [7,8]. While there are several reports of structured M&M conferences in HICs and a few in LMICs, there are no published reports of collaborative efforts to define a regional standard for M&M conferences with an accompanying toolkit [9,10].
This project aimed to establish regional (Lima, Peru) consensus on a standard for surgical M&M conferences and facilitate diffusion of this standard via a practical, simple M&M ''toolkit.'' To determine the effects of passive diffusion of the toolkit by stakeholders in Lima (Peru's capital city), we conducted serial evaluation through anonymous questionnaires to assess changes in M&M conferences practice in Peru prior to, and three years following, toolkit development.

Materials and methods
This project represents ''phase two'' of a multi-phase project to implement quality M&M conferences in Peru. Phase one included a baseline assessment of existing practices in Peru and Latin America and a review of global implementation of the WHO TQIP guidelines [11][12][13]. It revealed standardized case selection criteria, documentation of M&M conclusions and a clear plan for follow-up to predict M&M conferences which are perceived to result in institutional change [11][12][13].
The second phase, presented in this manuscript, reflects an initial response to the data collected: development of a ''toolkit'' designed to be concise, self-explanatory and applicable to a diverse array of hospital and service types.

Toolkit: development and dissemination
The Quality Chapter of the Peruvian General Surgery Society convened a working group. Individuals identified for the working group included those interested in quality improvement (QI) and having prior involvement with the Quality Chapter. Furthermore, the group was devised to include one or two general surgeons from each of the ten largest public, social security or military (non-pediatric) hospitals in Lima. Input was also provided by outside experts in trauma QI.
At the initial meeting, a summary of the relevant data from both the systematic review and the quantitative and qualitative assessment of QI practices in Latin America was presented to group participants. The group was then asked to reflect on these data and offer additional insights into particular areas for improvement in M&M conference practices in Lima. The working group met periodically over a six-week period. During this time, several of the working group members attended a course on trauma QI developed by the Panamerican Trauma Society.
The working group developed draft M&M toolkit based on examples from recent literature [9,10]. This was revised iteratively. Working group member participants trialed an interim version of the toolkit components in their institutions. Their feedback was incorporated into its final revision. The final version was submitted to the Peruvian General Surgery Society leadership for ratification and dissemination.
Initial dissemination included passive diffusion by key stakeholders at the ten included hospitals in Lima. Three years after toolkit development, we conducted an interval assessment of TQIPs status in Lima.

M&M conferences: assessment
The effect of the toolkit was evaluated by comparing the results of surveys on trauma quality improvement practices conducted in 2016 (before toolkit dissemination) and 2019. Baseline data (2016) for comparison came from a previously published survey [12]. This survey was repeated in 2019. In each case, the same questionnaire was used, the same interview procedures followed, and similar types of participants responded. We revisited nine out of 10 hospitals surveyed in 2016. The questionnaire included respondent demographics, hospital descriptors, self-reported QI practices at the respondent's hospital and M&M conference characteristics. Respondents included surgeons, residents and medical students. Respondents were contacted by research assistants at the Peruvian General Surgery Society, at other conferences and at hospitals in Lima.
Results from 2016 and 2019 were compared. The R project for statistical computing (Statistical Computing, Vienna, Austria) was used for data analysis. We tested for significance using Fisher's exact tests since all responses were categorical. Unadjusted p-values are presented together with p-values adjusted for multiple comparisons using Holm's method [14].
Multivariable logistic regression was employed to assess which variables were associated with a reported institutional change resulting from M&M conferences. We included M&M conference frequency, whether the respondent reported: ''improve the system'' as the main objective of M&M conferences, perceived most valid source of medical knowledge, presence of standardized case selection criteria, absence of barriers to referring cases, note-taking during M&M conference, the proportion of case presentations missing information, number of attendings present, follow-up plan, opportunity to discuss errors, presence of trauma registry, survey year, and whether lack of interest or lack of staff education were listed as primary obstacles. ''Number of M&M conferences per year'' was the only variable not coded as dichotomous; these were derived from the question ''How often do M&M conferences occur.'' We employed a series of one-variable logistic regression models to evaluate each covariate's unadjusted relationships and reported change. It is important to mention that the definition of ''system'' used during the development and dissemination of the toolkit was the one used by WHO [5], which was also aligned with the concept that local surgeons had about system. They understood ''system'' as in hospital surgical care of patients. Excluding prehospital care and hospital transport (both were not referenced in toolkit components) as these varies across Peru's different regions, mainly because Peru lacks a unified and standardized trauma care, especially in the prehospital setting. Furthermore, during the toolkit development meetings there was a consistent education regarding M&M conference to be more about ''system improvement'' rather than blaming someone for a patient's outcomes.

Results
The following four toolkit components were developed by the working group.

Toolkit component #1: guide for planning M&M conferences
The first toolkit component was developed to provide a guide, or checklist, to assist in planning a new or improving an existing M&M conference. This single-page document included conference logistics and case selection, presentation, discussion, documentation and follow-up. The guide includes evidence-based recommendations regarding M&M conference best practices and provides various options to allow the planner to modify the M&M processes according to what best serves their clinical and institutional scenario (Fig. 1).

Toolkit component #2: form for documenting M&M conferences
This single-page document is designed to provide a record of the M&M case discussion to facilitate monitoring of trends in mortality and complications, to prompt conference participants to complete all essential components of an M&M discussion and to provide a place to document planned corrective actions to facilitate accountability and follow-up of those actions. The form included space to document anonymous patient description (age, sex, primary diagnosis), case selection criteria, an assessment of the preventability of the death or complication, an assessment of the primary cause of the death or complication, a detailed description of up to two planned corrective actions with a written designation of the person responsible for those actions and the timeline for completion (Fig. 2).  [9]. It assists the junior clinician, often a resident, in selecting what information is salient in developing a concise case presentation, which includes all information necessary to inform a discussion of preventability and root cause. During the iterative toolkit development process, it was determined that the assessment and recommendations would be better left out of the presentation and included only in the discussion. Participants felt that residents did not have adequate experience with QI programs, root cause analysis and M&M conferences to make preliminary suggestions regarding assessment and recommendations. Furthermore, working group participants felt that the healthy interpersonal relations of the group would be more fully preserved if residents were not put in the position of offering assessment and recommendations for those senior to them (Fig. 3).

Toolkit component #4: code of conduct
Phase one research suggested that interpersonal conflict caused many failed attempts or non-sustainable M&M conferences in Lima. One suggested response to this common problem was the development of a ''code of conduct'' which would be signed by all participants and posted in a visible, central location (i.e., the chief's office or the conference room). This code included participant commitment to mutual respect, openness to opinions of all conference participants, confidentiality and a focus on system improvement rather than searching for blame (Fig. 4).

Interval assessment of TQIPs in Peru
A cross-sectional survey with 105 responses was conducted in 2019 and compared to the survey with 101 responses, which had been collected in 2016, three years prior to pretoolkit development. Data from the two time periods reflected sampling of groups with similar demographics, but with fewer respondents from private hospitals in 2019, as compared to 2016 (16.1% vs. 2.1%, p = 0.02) and a higher proportion of respondents were medical students in 2019 as compared to 2016 (26.5% vs. 6.3%, p = 0.01) ( Table 1). M&M conference characteristics were similar between the two time points, except for a trend toward lower reported attendance (Table 2). Additionally, there were more respondents in 2019 reporting ''improve the system'' as a perceived objective compared to 2016 (70.5% vs. 38.6%; p \ 0.001) ( Table 3).
Significantly more medical students participated in 2019 compared to 2016. However, when data were analyzed only among attendings, no statistically significant differences were identified between the two time periods. Table 4, a multivariable model of the 2019 and 2016 Peruvian data, shows the only significant predictor of whether a respondent reported the presence of institutional change attributable to an M&M conference to be the presence of a trauma registry (p = 0.05) as compared to 2016 regression model [12]. This was associated with a decrease in the odds of reported change by a factor of 0.40 (95% CI: 0.16-0.97) (Fig. 5).

Discussion
We sought to establish a regional standard for surgical and emergency M&M conferences and develop a simple toolkit to facilitate dissemination of that standard. The toolkit developed has significant overlap with those based at single institutions in high-income countries. However, the toolkit is distinct in its brevity-consisting of three single-page documents and a PowerPoint template, level of detail and comprehensiveness, self-explanatory and user-friendly format, and formation based on a consensus from individuals from more than eleven institutions and two continents.
This represents phase two of a three-phase endeavorand is a response to the baseline assessment conducted in phase one. This toolkit aimed to address the three components associated with an effective M&M conference in previous studies [9]. These three components include modifiable but specific case selection criterion-included on both the planning guide and case documentation form; the case documentation form itself; and inclusion on the case documentation form of a written plan for follow-up. Phase three, and the next steps in this project, is the structured dissemination and implementation of the toolkit, with subsequent evaluation of provider perception of educational value, collegial environment of the M&M conference and its effect on patient outcomes. This toolkit may serve as a basis for similar processes in other countries-by which a national organization engages a group of stakeholders to modify the toolkit to fit the local culture and any  Tables 1, 2 and 3), the latter of which may be explained by the increased number of medical students included in the surveys. No evidence of a change over these three years was found when analyzing only attending responses. However, the trend toward increased reporting of a focus on system-level changes persisted, but did not reach the level ''significance'' because as a subsample it has less power to find evidence against the null hypothesis.
The decrease in physicians' attendance percent might also reflect a heightened standard for M&M conferences definition after the development of the M&M toolkit and exposure to these QI endeavors. In the multivariable regression model, the estimated odds ratio for trauma differed from resulted published in 2016 [12]. This might be explained by the fact that only Peruvian data was used for this regression model, thus making our sample smaller than in 2016 (which included other countries from the Andean region). Therefore, current sample had less power to reach statistical significance.

Limitations
The utility of the toolkit itself may be limited by several factors. First, due to the logistical challenges of convening persons from broad geographic areas, the working group was comprised entirely of urban physicians. They worked at large academic and non-academic, public and private, institutions. Nonetheless, it may be that the toolkit has not addressed particular issues which would be relevant to a rural provider.
Additionally, this may also be affected by respondents being aware of research team observation as most of the surveys were complete in their presence (Hawthorne effect). Further, we used convenience sampling that may impact results.
This evaluation included responses from people who worked at a wide variety of institutions in Peru. The potential effect of the toolkit at the involved hospitals might be obscured by responses from hospitals to which the toolkit was not yet disseminated. Finally, it also important to highlight that there was not clear definition of ''system'' included in the survey, therefore that gave respondents room for interpretation, however, in a context familiar with ''system'' definition according to WHO guidelines and M&M conferences were more educational rather than punitive.

Conclusions
A toolkit was developed to make trauma M&M conferences more standardized and effective. This was created with the collaboration of surgeons from the 10 major hospitals in the largest city in Peru. A general evaluation of trauma QI programs three years after this revealed minimal changes, except for an increase in the percent of respondents who felt that improving the system was the objective of M&M conferences. Next steps should be an active dissemination of the toolkit and a more in-depth evaluation of the extent its utilization. We recommend increased national-level prioritization of high-quality M&M conferences as a building block of TQIPs in Peru.
Funding Grant: The project components were supported by the National Institutes of Health (NIH) through Fogarty International Center of the NIH under grant D43 TW009345 awarded to the Northern Pacific Global Health Fellows Program.

Declarations
Conflict of interest The authors declare that they have no conflict of interest.
Informed consent This project was reviewed by both the University of Washington and Universidad Peruana Cayetano Heredia Institutional Review Boards and was considered exempt. The research team asked participants for verbal consent and provided an informational sheet about the study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.