Device failures and patient adverse events from EUS: 26-year FDA MAUDE analysis (2000–2025)

Abstract

Background and Objectives:

EUS has emerged as an essential instrument in the field of gastroenterology, significantly enhancing both diagnostic evaluations and therapeutic procedures. By delivering high-resolution imaging and precise guidance, EUS facilitates a range of interventional techniques. This study sought to examine the adverse events linked to EUS devices following their approval by the US Food and Drug Administration.

Methods:

We performed a retrospective analysis utilizing postmarketing surveillance data sourced from the Food and Drug Administration’s Manufacturer and User Facility Device Experience (MAUDE) database. This analysis encompasses data from January 1, 2000, to August 31, 2025.

Results:

A total of 7608 reports detailing 8778 device-related issues and 7875 patient complications were analyzed. The most commonly reported device-related problems included peeling or delamination, with 1251 reports, followed by device breakage at 837 reports, microbial contamination at 781 reports, and inadequate cleaning at 704 reports. In terms of patient-related complications, the most frequently reported adverse events were blood loss (73 reports), tissue perforation (44 reports), and pancreatitis (32 reports). Furthermore, there were 743 documented instances of microbial contamination in EUS devices, with Coagulase-negative staphylococci being identified as the predominant pathogen, appearing in 151 reports.

Conclusions:

The examination of the MAUDE database reveals significant device and patient-related adverse events that endoscopists must consider during EUS procedures. Continuous postmarketing surveillance and the evaluation of real-world data are crucial for determining the safety and effectiveness of endoscopic devices, which in turn enhances patient outcomes.

INTRODUCTION

The EUS has transformed from a diagnostic tool into a crucial platform for advanced therapeutic interventions. Over the past decade, the use of EUS has significantly broadened, especially in the diagnosis and management of pancreatic and biliary disorders.^[1] EUS-guided tissue acquisition has emerged as the preferred technique for diagnosing various lesions. In addition to its diagnostic capabilities, it now plays a significant role in therapeutic procedures, including the drainage of pancreatic pseudocysts and walled-off necrosis. This approach also facilitates alternative biliary access and drainage, enables vascular interventions, and allows for targeted tumor ablation.^[2–6]

The increasing clinical importance of EUS is largely due to continuous improvements in equipment technology. Initial EUS systems provided basic cross-sectional ultrasound imaging, which, while effective, suffered from resolution limitations and required intricate operational skills. In contrast, contemporary EUS devices include a variety of advanced features. Currently, 4 types of instruments are utilized in clinical practice: radial scanning endoscopes, linear endoscopes, forward-view endoscopes, and catheter probes (miniprobe).^[7] Each of these devices has unique technical characteristics and specific clinical applications. The leading manufacturers of currently available EUS equipment include Olympus (Olympus Medical Systems Corporation, Center Valley, PA, USA), Pentax Medical (Montvale, NJ, USA), and Fujifilm Medical Systems (USA Inc., Lexington, MA, USA).

While EUS is typically regarded as a safe procedure, its expanding use as a therapeutic tool has resulted in a heightened risk profile. Adverse events now encompass not only common endoscopic issues, such as throat irritation and bloating, but also more serious complications. These complications include bleeding, infection, acute pancreatitis, perforation, and, although rare, death, especially during procedures involving tissue sampling, targeted punctures, and other therapeutic interventions.^[8] As EUS technology progresses, the incidence of technical issues and equipment failures has increased. Problems like needle breakage, probe damage, or malfunctioning operating channels can disrupt procedures and compromise patient safety.^[9]

The US Food and Drug Administration’s (FDA’s) Manufacturer and User Facility Device Experience (MAUDE) database is a national repository for adverse events related to medical devices in real-world clinical settings. It provides valuable data for assessing the safety profiles of EUS devices.^[10] This study leverages the MAUDE database, spanning from January 1, 2000, to August 31, 2025, to investigate adverse events related to both patients and devices in EUS. The main objective is to identify trends and patterns in significant adverse events, evaluate the safety profiles of different EUS devices, identify factors contributing to severe injuries, and formulate evidence-based recommendations aimed at improving patient safety.

METHODS

Data source and study design

We performed a retrospective analysis of postmarketing surveillance data sourced from the US FDA’s Manufacturer and User Facility Device Experience (MAUDE) database. The main aim of this study was to identify failure modes, patient-related injuries, and mortality associated with EUS devices. For our analysis, we downloaded the entire database, which includes all medical device reports from January 1, 2000, to August 31, 2025, on October 3, 2025. The MAUDE database serves as a publicly accessible, passive surveillance system that compiles both voluntary and mandatory reports of device-related adverse events from a variety of sources, including patients, consumers, healthcare providers, manufacturers, device user facilities, and importers. Updated monthly, this database is freely accessible to the public (https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/search.cfm).

Search strategy and device identification

A thorough search strategy was implemented to locate reports concerning EUS devices. Due to the absence of a distinct device category for EUS in the MAUDE database, we adopted a free-text search approach targeting the “GENERIC_NAME” field to enhance sensitivity. This search included a comprehensive set of 28 search terms [Supplementary Table 1, https://links.lww.com/ENUS/A395], applied in a case-insensitive manner and utilizing wildcards to capture morphological variations.

Data cleaning and screening process

The initial search identified a total of 8160 unique reports categorized by “MDR_REPORT_KEY.” A Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)-style flow diagram (Supplementary Figure 1, https://links.lww.com/ENUS/A395) has been included to visually represent the entire process of record identification, screening, eligibility assessment, and inclusion. This diagram outlines the retrieval of 8160 reports, the elimination of 141 duplicates, and the screening of 8019 reports. It also highlights the exclusion of 407 reports that were performed by non-EUS-specialized departments or pertained to non-EUS-specific devices, along with 4 reports that were missing critical data. Ultimately, 7608 reports were included in the final analysis.

Data structure and duplicate handling

The MAUDE database facilitates the documentation of multiple device-related issues and patient complications within a single report. Consequently, this study focuses on individual events rather than the reports themselves. This distinction clarifies why the total counts of device problems (n = 8778) and patient complications (n = 7875) surpass the total number of unique reports (n = 7608). To address the potential for duplicate reporting of the same clinical incident by various sources, our main deduplication approach utilized the unique “MDR_REPORT_KEY” identifier found in the database. Reports that shared the same “MDR_REPORT_KEY” were classified as duplicates and merged into a single entry, following the outlined screening process. However, it is crucial to recognize that independent reports from different sources regarding the same event may not have the same “MDR_REPORT_KEY,” which means they could still be included in the dataset. A comprehensive deduplication process based on event date, device serial number, and narrative similarity proved impractical due to the extensive volume of data and frequent gaps in these fields.

To evaluate the potential implications of undetected duplicates on our results, we conducted a sensitivity analysis using a subset of reports submitted solely by manufacturers (n = 3657). The distribution and ranking of the most common device problems and patient complications in this manufacturer-exclusive subset closely mirrored the findings from the complete dataset [Supplementary Tables 2 and 3, https://links.lww.com/ENUS/A395]. Although minor discrepancies in absolute frequencies were noted, the overarching patterns and identification of the most prevalent issues remained highly consistent. This strong degree of concordance supports the robustness of our primary conclusions, even in light of the possibility of residual duplication within the full dataset.

Data analysis and categorization

Each report was meticulously examined to extract key details, such as the event date, specifics of the device malfunction, and associated patient outcomes. Within the MAUDE database, adverse events are classified into 4 main categories: injury, death, device malfunction, and other. Our analysis encompassed all these categories. Device models were classified solely based on the “MODEL_NUMBER” field. We conducted a quantitative frequency analysis of all reported models. To enhance the reliability of our findings and avoid overinterpreting infrequent events, we set a threshold of 10 reports. Models that met or surpassed this threshold were individually listed in Table 1, whereas those with fewer reports were aggregated into an “other” category. All reports were included in the overall analyses. To quantitatively assess temporal trends in adverse event reporting, we employed Joinpoint regression analysis (Joinpoint Regression Program, Version 5.4.0, National Cancer Institute, Bethesda, MD, USA) on the annual count of unique reports from 2003 to 2025. This method identifies significant shifts in trends, known as joinpoints, and calculates the annual percent change (APC) for each segment, testing whether the APC significantly differs from zero. The optimal number of joinpoints was established using the Bayesian Information Criterion. To ensure clarity and reproducibility in our categorization process, we provide the operational definitions and adjudication rules that link the source MAUDE codes to our final clinical impact categories in Supplementary Table 4, https://links.lww.com/ENUS/A395.

Table 1.

Summary of EUS models.

Company	Events, n (%)	Device issue(s)	Type	Events, n (%)
Olympus Medical Systems Corporation	5536 (72.77%)	GF-UCT180	Linear ultrasound endoscope	3529 (46.39%)
		GF-UCT260	Linear ultrasound endoscope	948 (12.46%)
		GF-UE160-AL5	Radial ultrasound endoscope	586 (7.70%)
		GF-UC140P-AL5	Linear ultrasound endoscope	111 (1.46%)
		GF-UE260-AL5	Radial ultrasound endoscope	111 (1.46%)
		GF-UE190	Radial ultrasound endoscope	89 (1.17%)
		GF-UCT140-AL5	Linear ultrasound endoscope	78 (1.03%)
		GF-UE290	Radial ultrasound endoscope	33 (0.43%)
		GF-UC140P-AL5 & GF-UCT180	Linear ultrasound endoscope	21 (0.28%)
		TGF-UC260J	Linear ultrasound endoscope	18 (0.24%)
		GF-UCT240-AL5	Linear ultrasound endoscope	12 (0.16%)
Hoya Corporation Pentax Tokyo Office (Pentax of America, Inc)	1957 (25.72%)	EG-3870UTK	Linear ultrasound endoscope	594 (7.81%)
		EG38-J10UT	Linear ultrasound endoscope	587 (7.72%)
		EG-3670URK	Radial ultrasound endoscope	285 (3.75%)
		EG36-J10UR	Radial ultrasound endoscope	217 (2.85%)
		EG-3270UK	Linear ultrasound endoscope	190 (2.50%)
		EG34-J10U	Linear ultrasound endoscope	84 (1.10%)
Other (n of 10 or less)	115 (1.51%)	-	-	-

Ethical considerations and limitations

This study relied on publicly available data that did not include any protected health information, thus eliminating the need for institutional review board approval. The research was conducted without any direct patient engagement, financial costs, or associated risks. It is worth noting that although this surveillance system does not provide exact event rates due to the lack of publicly accessible data regarding overall device usage, it does provide significant insights into prevalent challenges and potential issues concerning medical devices.

RESULTS

We retrieved a total of 7608 reports from the FDA’s MAUDE database, which documented 8778 device-related issues and 7875 patient complications associated with EUS from January 1, 2000, to August 31, 2025. The disparity between the number of reports and events arises because a single report can encompass multiple device problems and/or patient complications. Joinpoint regression analysis of annual report counts starting from 2003 indicated a significant upward trend, with the APC reaching 68.24 (P < 0.01), culminating in a peak of 2053 reports in 2023. This upward trend was followed by a significant decreasing trend from 2023 to 2025 (APC = −51.35, P = 0.01) [Supplementary Figure 2 and Supplementary Table 5, https://links.lww.com/ENUS/A395]. The earliest report dates back to 2003, whereas 2023 marked the highest number of reports (2053 reports) [Figure 1 and Supplementary Table 6, https://links.lww.com/ENUS/A395]. We infer that the decrease in reports during 2025 may be influenced by the data collection period ending in August of that year.

Figure 1.

Annual trends in unique reports of EUS–associated adverse events from the MAUDE database (2003–2025). This graph shows the annual number of unique medical device reports (MDRs) associated with EUS devices. The y-axis indicates the number of reports, and the x-axis shows the calendar years. The unit of analysis is a unique report, consolidated by the MDR_REPORT_KEY identifier to prevent duplication of the same report. Data for 2025 are partial and include reports from January 1 to August 31, 2025. The total number of unique reports across the entire study period is 7608.

Device failures

A total of 8778 device-related issues involving EUS were identified and classified into 5 primary categories: material and structural problems (3857reports, 43.94%), cleaning and contamination issues (2115 reports, 24.09%), fluid channel and flow abnormalities (1102 reports, 12.55%), display and optical failures (828 reports, 9.43%), and component and assembly defects (272 reports, 3.10%) [Table 2]. The most frequently reported device-related issues were peeling or delamination of components (1251 reports, 14.25%), device breakage (837 reports, 9.54%), microbial contamination (781 reports, 8.90%), and inadequate cleaning (704 reports, 8.02%) [Figure 2 and Supplementary Table 7, https://links.lww.com/ENUS/A395].

Table 2.

Summary of EUS device problems.

Major classification	Events, n (%)
Material and structural issues	3857 (43.94%)
Cleaning and contamination issues	2115 (24.09%)
Fluid channel and flow abnormalities	1102 (12.55%)
Display and optical failures	828 (9.43%)
Component and assembly defects	272 (3.10%)
Operation and functional failures	213 (2.43%)
Other issues	391 (4.45%)

Figure 2.

Temporal trends in EUS device problems: absolute counts and relative proportions. A, Stacked bar chart of annual EUS device problem reports. The y-axis represents the number of unique reports per year, and the x-axis shows the calendar years. The height of each bar indicates the total number of reports for that year, with colored segments representing the count of reports attributed to each of the 7 major problem classifications (defined in the legend). A single report can document multiple device problems. B, Proportional distribution of EUS device problem categories. This stacked area chart illustrates the relative percentage (y-axis, from 0% to 100%) of reports attributed to each of the 7 problem classifications for each year (x-axis). Problem classifications are defined as follows: Material and structural issues: problems related to the physical integrity of device components, including peeling/delamination, breakage, cracks, tears, and connection failures. Cleaning and contamination issues: failures in reprocessing, including inadequate cleaning, chemical residue, and microbial contamination of the device. Fluid channel and flow abnormalities: issues affecting irrigation and suction, including leaks, blockages, and flow control failures. Display and optical failures: problems with the video and imaging system, including loss of display, image artifacts, and optical obstructions. Component and assembly defects: defects related to missing, loose, or improperly assembled components. Operation and functional failures: malfunctions affecting device operation, including mechanical jams, unintended movement, and output abnormalities. Other issues: all other reported problems, including incomplete information, scratches, and usage difficulties. Data were retrieved from the MAUDE database. Data for the year 2025 are partial, covering the period up to August 31, 2025. The total number of device problem events analyzed is 8778.

Reports of problems related to EUS devices were relatively low before 2020; however, a significant increase was observed starting in that year, peaking between 2022 and 2023, followed by a slight decline in 2024 (Figure 2A). Among the various categories, material and structural issues experienced the most significant rise, accumulating 2935 reports from 2020 to 2023. Cleaning and contamination issues were the second most reported, with 1365 cases during the same period [Figure 2 and Supplementary Table 8, https://links.lww.com/ENUS/A395].

Device-associated patient adverse events

A total of 7875 patient complications were associated with EUS, which were categorized into 11 groups: no clinical impact (7108 reports, 90.26%), device/technical problems (276 reports, 3.50%), infectious complications (113 reports, 1.43%), tissue and organ injury (95 reports, 1.21%), hemorrhagic and hematologic issues (92 reports, 1.17%), unspecified events (76 reports, 0.97%), digestive disorders (48 reports, 0.61%), symptomatic manifestations (35 reports, 0.44%), respiratory problems (14 reports, 0.18%), cardiovascular concerns (11 reports, 0.14%), and life-threatening conditions (7 reports, 0.09%) [Table 3]. The most commonly reported adverse events were blood loss (73 reports, 0.93%), tissue perforation (44 reports, 0.56%), and pancreatitis (32 reports, 0.41%) [Supplementary Table 9, https://links.lww.com/ENUS/A395]. The majority of complications associated with EUS were events that did not have any clinical significance, making up a substantial proportion compared with other complication categories [Supplementary Figure 3 and Supplementary Table 10, https://links.lww.com/ENUS/A395]. This finding highlights the overall safety profile of EUS procedures. Among the infectious complications reported, 3 distinct types were identified: infections affecting the digestive system (63 reports, 0.8%), systemic infections (35 reports, 0.44%), and localized infections (15 reports, 0.19%) [Figure 3 and Supplementary Table 11, https://links.lww.com/ENUS/A395].

Table 3.

Summary of patient adverse events.

Major classification	Events, n (%)
No clinical impact	7108 (90.26%)
Device/technical abnormalities	276 (3.50%)
Infectious complications	113 (1.43%)
Tissue and organ injury	95 (1.21%)
Hemorrhagic and hematological issues	92 (1.17%)
Unclassified events	76 (0.97%)
Digestive system issues	48 (0.61%)
Symptomatic manifestations	35 (0.44%)
Respiratory system issues	14 (0.18%)
Cardiovascular system issues	11 (0.14%)
Life-threatening conditions	7 (0.09%)

Figure 3.

Temporal trends in EUS–associated infectious complications: annual counts and subtype distribution. A, Annual number of infectious complication events. The bar plot depicts the total number of infectious complication events reported per year. The y-axis represents the number of events, and the x-axis shows the calendar years. The unit of analysis is the infectious complication event. A single report can contribute multiple events if it documents more than 1 type of infectious complication (e.g., both pancreatitis and an abscess). B, Proportional distribution of infectious complication subtypes. The stacked area chart illustrates the relative percentage (y-axis) of events attributed to each of the 3 predefined infectious complication categories for each year (x-axis). The proportions were calculated based on the total number of infectious complication events per year. Infectious complication categories are defined as follows: Digestive system infections: includes pancreatitis, cholangitis, and bacterial/fungal peritonitis. Systemic infections: includes septicemia, septic shock, and nonspecific systemic bacterial infections. Localized infections: includes abscess formation and aspiration pneumonia. Data are derived from the MAUDE database. Data for the year 2025 are partial, covering the period up to August 31, 2025. The total number of infectious complication events across the study period is 113.

Identified microbiologic contamination

A total of 743 reports of microbial contamination in EUS were recorded, with 601 cases identifying specific microorganisms through bacterial culture. The most commonly reported pathogens included Coagulase-negative staphylococci (151 reports, 20.32%), Pseudomonas aeruginosa (103 reports, 13.86%), and Micrococcus spp. (82 reports, 11.04%), which together accounted for 55.91% of all bacterial infection reports (336 out of 601) [Supplementary Table 12, https://links.lww.com/ENUS/A395].

Manufacturer and model distribution of reported EUS devices

Olympus Medical Systems Corporation devices were the most frequently reported, accounting for 5536 reports (72.77%). It is essential to interpret these figures carefully, as they primarily indicate market exposure, sales volume, and reporting behaviors rather than serving as a measure of the safety or performance of different manufacturers or models. Notably, the GF-UCT180 linear ultrasound endoscope emerged as the most cited model, accounting for 3529 reports (46.39%). Devices from Hoya Corporation Pentax contributed 1957 reports (25.72%), primarily involving the EG-3870UTK (594 reports, 7.81%) and EG38-J10UT (587 reports, 7.72%) models. Linear ultrasound endoscopes were the primary devices involved in adverse events across all manufacturers. The Olympus GF-UCT180 accounted for nearly half (46.39%) of the model-specific reports, likely due to its market dominance and the technical complexity involved in the interventional procedures conducted with this device.

DISCUSSION

Our extensive examination of the FDA’s MAUDE database, spanning 26 years, uncovered notable trends in device-related problems and their associated patient complications in EUS. From 2020 to 2023, we observed a significant rise in device-related issues, particularly concerning structural defects and component delamination. Nonetheless, it is important to highlight that an overwhelming majority of reported incidents (90.26%) did not have any clinical complications; however, instances of more serious complications, including bleeding, perforation, and pancreatitis, were observed. Microbial contamination was most commonly linked to Coagulase-negative staphylococci, P aeruginosa, and Micrococcus spp. Although EUS procedures are generally safe, the identified structural issues, infection trends, and emerging patterns emphasize the need for targeted monitoring and improvements in clinical practices.

Device-related problems: structural integrity and contamination

Our study revealed that material and structural defects are the primary issues, comprising 43.94% (3857 reports) of all documented problems. The second most common concern was cleaning and contamination, with 2115 reports, representing 24.09% of the total. Among the structural defects, the most frequently reported issues were delaminated or detached components, which accounted for 1251 reports (14.25%). Other significant issues included device breakage (837 reports, 9.54%) and tears, splits, or cuts in the material (461 reports, 5.25%). These structural failures not only lead to immediate procedural difficulties but also pose significant long-term risks to patient safety.^[11,12] Structural damage can create gaps of varying sizes that provide a suitable environment conducive to pathogens. These pathogens may endure conventional cleaning methods, thereby elevating the risk of infection, even when some damages go unnoticed.^[13] Compromised structural integrity can disrupt the transmission and reception of ultrasound signals, diminishing image quality. This decrease in precision could contribute to diagnostic inaccuracies and increase the risk of unintended injuries during procedures like fine-needle aspiration.^[14]

Patient adverse events and infectious complications

Current guidelines emphasize the critical need to prevent contamination; however, the integrity of medical devices often receives insufficient attention. This oversight may inadvertently lead to procedural complications and facilitate the spread of infections. Our research underscores the necessity for regular inspections to identify early signs of structural damage, thereby preventing potential clinical issues. Although device-related problems are not uncommon, our analysis reveals that 90.26% (7108/7875) of reported patient complications had no clinical impact, suggesting that EUS procedures are generally safe. In instances where clinical consequences were observed, the most frequently reported severe complications included blood loss (73 reports), tissue perforation (44 reports), and pancreatitis (32 reports). These results align with previous studies on complications associated with EUS-guided fine-needle aspiration, which similarly identified bleeding and mechanical injuries as leading risks.^[15]

Infectious complications were documented in 113 reports (1.43%) and were classified into 3 groups: digestive system infections (63 reports, 0.8%), systemic infections (35 reports, 0.44%), and localized infections (15 reports, 0.19%). The majority of these infections were digestive system-related, with pancreatitis (32 reports, 0.41%) being the most frequently observed. This distribution underscores the inherent risks associated with EUS procedures, particularly those that necessitate transmural access. Within the category of systemic infections, septic shock (7 reports, 0.09%) and septicemia (5 reports, 0.06%) were the most severe, suggesting that localized infections can escalate into more severe health issues if they are not promptly recognized and treated effectively. Additionally, there were a total of 743 instances of microbial contamination in EUS devices, with 601 cases involving identified pathogens.

The most commonly identified organisms included Coagulase-negative staphylococci (151 reports, 20.32%), P aeruginosa (103 reports, 13.86%), and Micrococcus spp. (82 reports, 11.04%). Together, these pathogens accounted for 55.91% (336/601) of all confirmed bacterial infections. The clinical significance of these findings lies in the unique challenges each pathogen presents. The frequent presence of Coagulase-negative staphylococci, typically found on skin, suggests contamination due to improper handling or human contact during the cleaning process.^[16] P aeruginosa is of particular concern due to its ability to form biofilms and its resistance to many disinfectants. Studies have demonstrated that this pathogen can persist in small crevices of endoscopes, even after standard high-level disinfection protocols.^[17] The prevalence of Micrococcus spp., resilient environmental and skin commensals, highlights the persistent challenges associated with device reprocessing. These microorganisms are adept at surviving in dry environments, which may enable them to remain on equipment surfaces and potentially lead to device-related infections.

The pattern of infectious complications identified in our study appears to be influenced by the evolving clinical use of EUS, especially as it becomes increasingly integrated into complex surgical procedures. Although our data do not specify the procedural intent for all cases reported, the severity of the infections highlights a likely connection to transmural puncture and drainage interventions. This finding aligns with the established higher risk associated with interventional EUS as opposed to solely diagnostic examinations. The transition of EUS into a primary therapeutic modality, especially for managing malignant double obstruction or biliary-pancreatic diseases in patients with surgically altered anatomy, inherently involves breaches of mucosal barriers and the potential for microbial inoculation.^[18,19] Therefore, the shift towards a more operative echoendoscopic practice necessitates heightened vigilance and reinforced infection control protocols to mitigate these risks.

Manufacturer and model distribution

Olympus devices were most frequently reported in adverse event incidents, comprising 5536 reports (72.77%). The GF-UCT180 linear ultrasound endoscope was the most commonly mentioned model, accounting for 3529 reports (46.39%). Pentax devices were reported in 1957 reports (25.72%), with the EG-3870UTK (594 reports, 7.81%) and EG38-J10UT (587 reports, 7.72%) models being the most frequently cited. The raw frequency of reports concerning specific manufacturers and models in the MAUDE database should be interpreted with caution. These figures largely represent market exposure and the tendencies of reporting rather than actual variations in device safety or performance. For instance, the frequent appearance of Olympus devices in these reports is more indicative of their substantial market presence than any underlying safety concerns. Industry data show that Olympus commands around 70% of the global market share in gastrointestinal endoscopy, underscoring its preeminent position within this sector.^[20] The elevated number of reports concerning models like the GF-UCT180 should not be misconstrued as indicative of a higher risk profile. Instead, this trend reflects their extensive clinical application, especially in intricate interventional procedures that inherently involve greater device stress and reporting tendencies. A significant proportion of adverse events has been associated with linear echoendoscopes from multiple manufacturers, which corresponds with their prevalent use in these complex interventions. Different device models exhibit notable variations in critical design aspects, including the dimensions of the working channel, flexibility of the tip, and imaging acquisition techniques. Such differences can affect both performance and maintenance needs in specific clinical settings.^[21] The American Society for Gastrointestinal Endoscopy (ASGE) quality standards highlight that differences in device characteristics can result in variations in durability and failure modes.^[22]

Temporal trends and potential contributing factors

Our findings reveal a notable temporal shift in the frequency of EUS-related adverse events. Joinpoint regression analysis identified a significant upward trend from 2003 to 2023 (APC = 68.24, P < 0.01), which peaked at 2053 reports in 2023, followed by a significant decrease from 2023 to 2025 (APC = −51.35, P = 0.01). We propose that the observed escalation is driven by multiple factors, aligning with various global events and specific developments within the healthcare sector. Although our data do not allow us to establish a direct causal relationship, the trend we have noted aligns with the increasing utilization of EUS, particularly for interventional procedures. This rise would naturally lead to greater device usage and a higher volume of reports. Additionally, the temporal overlap with the COVID-19 pandemic indicates it may have played a role in this trend. It is reasonable to suggest that the well-documented disruptions to healthcare services during this time, including challenges related to endoscope maintenance and reprocessing protocols, could have heightened existing risks.^[23]

Between 2020 and 2023, there was a significant rise in reports of material and structural issues, totaling 2935 incidents. We hypothesize that this increase may be attributed to several pandemic-related factors. These include the prolonged storage of devices during times of reduced elective procedures, the implementation of stricter and potentially more rigorous cleaning protocols, and the extensive use of powerful disinfectants, all of which have been anecdotally linked to accelerated device wear. Furthermore, supply chain disruptions compelled many facilities to rely on outdated equipment beyond the recommended service intervals. The notable surge in reports of peeled or delaminated components after 2020 may also stem from alterations in manufacturing processes, reprocessing methods, or a heightened awareness of these issues.

Additionally, staffing shortages during this timeframe likely hindered the thorough inspection and reprocessing of devices, as reflected in the increase in cleaning and contamination-related incidents (1365 reports). The spike in microbial contamination reports during this period likely reflects enhanced microbiological surveillance protocols, leading to more frequent detection of pathogens. These trends highlight the urgent need for continuous improvements in device design and maintenance practices, especially as the adoption of EUS technology continues to grow.

Implications for clinical practice and future directions

The evolution of EUS as a primary therapeutic approach necessitates a corresponding improvement in device management and safety protocols. Our analysis of real-world data identifies specific failure modes that can be addressed through targeted strategies.

First, it is essential to implement routine high-resolution integrity checks. Material peeling or delamination has emerged as the most frequently reported device issue, accounting for 14.25% of cases. Therefore, visual inspection protocols must go beyond basic functionality assessments. We recommend the regular use of borescopes or other magnification tools to scrutinize the distal end, bending section, and internal working channel for early indications of wear, cracks, or delamination. This is especially important for linear echoendoscopes, which endure heightened mechanical stress during interventional procedures.^[24]

Second, it is crucial to establish definitive triggers for preventive maintenance. Any identification of structural defects should necessitate the immediate removal of the device from clinical service. Such findings must initiate a thorough inspection and preventive maintenance conducted by qualified technicians. This proactive strategy is vital to prevent further damage, reduce the risk of biofilm formation in microscopic crevices, and ensure the ultrasound image quality remains optimal for accurate diagnosis and intervention.

Third, the verification process for reprocessing following device damage or repair must be strengthened. For any device that has been repaired due to structural issues, standard cleaning and high-level disinfection protocols should undergo heightened verification. We recommend implementing additional microbiological monitoring, such as microbial culturing or adenosine triphosphate bioluminescence testing, on the repaired device over several consecutive reprocessing cycles to ensure effective decontamination before it is returned to clinical use. Fourth, strengthen surveillance for high-concern pathogens. The high prevalence of Coagulase-negative staphylococci and P aeruginosa suggests specific vulnerabilities in handling and reprocessing.^[25] Endoscopy units should consider proactive monitoring, including periodic environmental cultures of reprocessing areas and random microbiological sampling of EUS devices, focusing on models with complex channel designs.

Future device designs must focus on utilizing materials and construction methods that maximize durability and prevent delamination. Additionally, advancements in reprocessing technologies, including automated channel flushing and systems for verifying drying processes, should be pursued and validated to more efficiently eradicate biofilm-forming pathogens such as P aeruginosa. By synchronizing clinical vigilance with technological innovation, the community can guarantee that the exceptional therapeutic advantages of EUS are provided while upholding the highest safety standards.

Limitations

This study acknowledges several limitations associated with utilizing the MAUDE database for postmarketing surveillance. First, the database operates as a passive reporting system, which is inherently prone to underreporting, incomplete narratives, and potential inaccuracies. Second, reporting biases can arise, as narratives submitted by manufacturers may differ from those provided by user facilities, potentially affecting the interpretation of the root causes of events. Additionally, around 50% of reports were missing the “REPORT_SOURCE” field, resulting in a high rate of missing data that hindered reliable stratified trend analysis and limited our analytical capabilities. The investigation of stratified temporal trends remains a crucial avenue for future research that could benefit from more comprehensive datasets.

Another significant challenge is the database’s structural separation of device contamination events from patient infectious complications, complicating efforts to establish direct microbiological causation between specific contamination incidents and resulting patient infections. A critical limitation for comparative assessments is the lack of denominator data, such as the number of devices in use or the total procedures performed for each model. Without this data, determining the true incidence or comparative risk of adverse events across different devices becomes impossible.

Furthermore, it is essential to note that our findings cannot be used to calculate incidence rates or true risk levels. This limitation arises from the absence of reliable denominator data and the potential inflation of the numerator (event counts) due to undetected duplicate reports and the aggregation of multiple device issues or patient complications within a single report. As a result, any comparisons of absolute report counts among manufacturers or models should be approached with caution and should not be interpreted as indicators of relative safety or performance. Our findings reflect counts and proportions within this specific dataset and should not be generalized to represent the overall safety profile of these devices. Despite these limitations, the MAUDE database continues to serve as a valuable tool for identifying device failure modes and detecting emerging safety signals.

CONCLUSIONS

Our extensive analysis of 26 years of data from the MAUDE database reveals important trends related to adverse events associated with EUS. We focus particularly on device integrity and microbial contamination, which are critical safety issues. Although most of the reported cases did not lead to serious clinical outcomes, the frequent occurrences of device malfunctions and contamination necessitate continued vigilance. In light of these findings, we recommend evidence-based strategies to improve device design, enhance reprocessing protocols, and strengthen safety measures. These steps are essential for promoting patient safety as EUS technology becomes more widely utilized in clinical practice.

Supplementary Information

Supplemental digital content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal's Web site (www.eusjournal.com).

Source of Funding

This work was supported by grants from the Science and Technology Commission of Shanghai Municipality (Grant No. 23Y11905300), Noncommunicable Chronic Diseases-National Science and Technology Major Project (Grant No. 2024ZD0520800), Fundamental Research Funds for the Central Universities (Grant No. YG2025ZD01), Shanghai Top-Priority Medical Center Construction Project (Grant No. 2023ZZ02022), National Natural Science Foundation of China (Grant No: 82570787), and Shanghai Hospital Development Center Foundation (Grant No: SHDC12024151).

Conflicts of Interest

The authors declare that they have no financial conflict of interest with regard to the content of this report.

Ethical Statements

This study is a retrospective analysis of publicly available, anonymized data from the U.S. FDA MAUDE database. As no private patient information was involved and no human subjects were directly recruited or interacted with, this research did not require review or approval by an Institutional Review Board (IRB) or Ethics Committee.

Author Contributions

S. Zhai, X. Lin, and Y. Song designed research; X. Lin, Y. Song, Z. Li, J. Chen, and X. Huang conducted research and data collection; Z. Ji, Q. Liu, Y. Xiao, and R. Han analyzed data; S. Zhai, R. Zeng, Y. Yang, W. Deng, Q. Zhang, and X. Lin wrote the paper. S. Zhai, W. Deng, and X. Li had primary responsibility for the final content. All authors read and approved the final manuscript.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available in the MAUDE repository, https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/search.cfm. Data described in the manuscript, code book, and analytic code will be made available upon request.