In low- and middle-income countries, the burden of poor-quality health care has not been adequately quantified due to data scarcity, lack of standardization, and insufficient research on quality interventions. Moreover, recent estimates suggest that in these countries, between 5.7 and 8.4 million people die annually due to adverse events (AEs), where an incident resulted in harm to a patient [1]. Among the common AEs that can lead to preventable harm to patients are medication errors, unsafe surgical procedures, health care–associated infections, diagnostic errors, patient falls, pressure injury, patient misidentification, unsafe blood transfusions, and venous thromboembolism [1,2]. Annually, the indirect costs of harm from unsafe health care practices amount to trillions of US dollars worldwide, with 4 out of every 100 patients in low-to-middle-income countries losing their lives due to substandard [3]. According to the World Health Organization (WHO), the financial and economic costs of safety lapses in the health care sector will be even greater and will remain a challenge for modern health care delivery in all countries. This is attributed to patient safety incidents resulting in fatalities, disabilities, and distress for both patients and their families, along with health care workers involved in serious patient incidents [2]. This situation reinforces the hypothesis that AEs are a significant yet neglected public health issue. Furthermore, during the 72nd World Health Assembly held in May 2019, the WHO recognized patient safety as a global health priority and an essential component for achieving universal health coverage across all countries. The WHO, in collaboration with Member States, adopted resolution WHA72.6 “Global Action on Patient Safety,” with the objective of transforming the movement into a social action where patients start demanding safer health care. This marks a strategic moment as this action has been globally acknowledged as a powerful policy tool that will shape the global patient safety agenda in the coming years and reduce the costs related to AEs at all levels of health systems [1].

There are several methods available for investigating and measuring AEs, such as voluntary reporting, retrospective review of medical records involving the analysis of past patient records, prospective analysis of the patient charts of inpatients which includes capturing real-time data, direct observation, interviews with patients and health care professionals, and analysis of indicators, reports, and complaints of poor practices [4]. Among these methods, one of the most used inpatient safety research is voluntary reporting, which relies on health care providers reporting AEs that they encounter during patient care [5,6]. Another frequently used methodology in research is manual and retrospective medical record review [5,7-10]. This method is considered the “gold standard” for estimating the occurrence of AEs in hospitalized patients due to its ability to capture detailed information and this involves trained reviewers examining patient charts to identify potential AEs, followed by physician review to determine if an AE indeed occurred [11].

In the year 2003, the Institute for Healthcare Improvement (IHI) developed the Global Trigger Tool for Measuring Adverse Events [8]. This approach involves external reviewers who gather information through triggers, which are clues or indicators recorded in archived medical records that signal the potential occurrence of an AE associated with patient care [7,8]. In 2001, the IHI proposed triggers focused on tracking AEs in the outpatient setting [8], known as the Outpatient Adverse Event Trigger Tool [12]. It is important to note that the first step of the patient safety research cycle proposed by the WHO involves measuring the magnitude of AEs [1].

Despite the advantages of trigger tools in AE tracking, the methodology presents disadvantages including variation in detected AEs [6,13], limited scope [7,10], resource-intensive usage [14-16], lack of specificity [16,17], delay in tool application [7,10,15,16], dependency on complete and legible documentation [7,18], reliance on human judgment [15,16], limited data extraction [15], need for data standardization [16], and limited sensitivity [7,10,18]. For these reasons, researchers have proposed the use of automated triggers to overcome the limitations of the manual method. In investigations conducted in the United States [14] and Switzerland [16,18], researchers used automated triggers to track AEs.

In another study, machine learning (ML) was used to assess risk factors associated with severe AEs, after tracking patient harm in a cancer hospital in Chongqing, China [19]. The findings indicated that automated triggers and ML successfully identified risk factors for severe AEs induced by antineoplastic drugs in patients with cancer. The ML model incorporated variables such as age, cancer type, number of prescribed medications, and other factors. The study’s outcomes were considered promising, as the developed model demonstrated the ability to predict severe AEs induced by antineoplastic drugs in patients with cancer. The researchers concluded that the use of automated triggers and ML is a promising approach to identifying risk factors for severe events, in contrast to the manual and retrospective method which aims to track events that have already occurred. Consequently, automated triggers represent a proactive approach, while triggers developed by the IHI are reactive.

Mixed methods research will be conducted within a design thinking framework. The study will take place at the outpatient unit of a prominent tertiary-level teaching hospital in the Metropolitan Region of Ribeirão Preto, located in the state of São Paulo, Brazil. The institution provides services across various medical specialties, offering exclusive services to the Unified Health System. Within the outpatient building is the institution’s largest surgical center, as well as the Central Laboratory Division—the first public laboratory in Brazil certified by the American College of Pathologists. The unit also has jointly a pharmaceutics unit, responsible for dispensing prescribed medications and regularly producing various categories of pharmaceuticals.

The study will be conducted in accordance with the Declaration of Helsinki and in compliance with the standards and guidelines established by the National Research Ethics Commission, as specified in the regulations and recommendations of Resolution 466/12 of the National Health Council [37], which regulates research involving human participants.

This project also involves documentary research as the incident or AE reports that occurred with ambulatory patients will be analyzed. It will not be possible to obtain informed consent from individuals whose data are contained in these restricted-access documents, as they do not attend the owning institution. In this case, the researcher will sign the Commitment Term for Data Use and Handling, which will be handled only after approval by the Research Ethics Committee and the owning institution.

The study was approved by the members of the organization within the institutions (146.00010241/2023-15) and it has been approved by the College of Nursing, Ribeirão Preto, University of São Paulo (Brazilian Ethics Board 6.811.085) and by the ethics board of the institution where the study will take place (Brazilian ethics board 6.851.281), enabling the commencement of data gathering in the ambulatory care unit.

Participants will be informed about the study’s objectives and, upon agreeing to participate voluntarily, they will be asked to sign the informed consent form. There will be 3 informed consent processes—2 separate informed consent for stages 1 and 2 (1 for patients, family members or caregivers, and another for participants who are members of the multidisciplinary health care team). The third informed consent will be considered for stage 4 participants in this study.

Participants will not receive direct benefits or compensatory benefits, but the results of this study may generate indirect benefits. The results will be useful for developing automated triggers in the outpatient setting. As mentioned in stage 1, under step 4, participants’ identification will be encrypted, to guarantee privacy and will not be disclosed when publishing the study results. All data obtained will be used for scientific purposes only.

In this paper, we present the research protocol in its preliminary stages, prior to any data gathering and analysis. This study is the first to create automated trigger tools in ambulatory settings in Brazil, presenting a valuable contribution to the health field while also adapting the design thinking framework. This framework is still underexplored in Brazilian research, especially in the context of patient safety. It is crucial to highlight the real-world application, specifically the effective implementation in the local realities.

In Brazil, despite a few published studies on the application of the trigger tool, most have focused on hospital settings [51,52], and more recently, on dental care [53]. However, according to the 2018 report from the Organization for Economic Cooperation and Development [54], about half of the global burden of AEs originated in primary health care, and 4 in 10 patients experience safety incidents, often resulting in hospitalizations and increased care needs. Moreover, the most common AEs in primary health care are related to medication use, largely caused by polypharmacy.

Given this context, in 2001, IHI proposed triggers for tracking AEs in the outpatient setting [12], known as the Outpatient Adverse Event Trigger Tool [55]. The tool comprises 11 triggers that provide clues or hints regarding the existence of an AEs in the patient’s medical record, including (1) new cancer diagnosis, (2) placement in a nursing home, (3) hospital admission and discharge, (4) 2 or more outpatient visits in a year, (5) surgical procedure, (6) emergency department visit, (7) use of more than 5 medications, (8) change of physician, (9) complaint letter, (10) more than 3 nursing calls in a week, and (11) abnormal laboratory value. These triggers were tested in American outpatient clinics but are certainly not the only possible triggers, although they represent an important starting point. According to IHI [8], each institution should choose to add, remove, or modify the triggers currently on the list to meet local realities. However, this study will develop automated triggers, in contrast to the manual and retrospective methods.

This study has the potential to positively impact the quality of care provided and simultaneously make a valuable contribution to the international scientific community. Thus, the best practices for outpatient care encompass principles of care continuity beyond hospital walls. In line with these observations, the importance and urgency of identifying events related to unsafe care in the outpatient context are emphasized to guide improvement and safety strategies for patients in these health care settings [56]. Furthermore, decisions can be made to minimize the burden of preventable harm, both economically and socially [57], as well as to promote a safety culture and establish high-reliability organizations.

This study will be conducted in an ambulatory setting with participants who are adults aged 18 years or older. It will exclude EMHR from pregnant individuals and psychiatric patients, which are anticipated limitations of this study.

For this study protocol, a separate paper focused on explaining the ML methods can be considered after the outcomes of stages 1 and 2 in this study. In the future, it will be used in every unit and will adjust the triggers based on each unit of the hospital. As mentioned, this is the first study to create automated triggers in Brazil. Therefore, there is a potential high demand for contributions nationally, and our study will also benefit other lines of research both nationally and internationally.

After the development of automated triggers in the outpatient setting, it will be possible to prevent and identify potential risks of AEs more promptly, providing valuable information. This technological innovation not only promotes advances in clinical practice but also contributes to the dissemination of techniques and knowledge related to patient safety. Additionally, health care professionals can adopt evidence-based preventive measures, reducing costs associated with AEs and hospital readmissions, enhancing productivity in outpatient care, and contributing to the safety, quality, and effectiveness of care provided. Additionally, in the future, if the outcome is successful, there is the potential to apply it in all units, as planned by the institutional organization.