Evaluation of reporting trends in the MAUDE Database: 1991 to 2022

Abstract

Adverse event reporting for medical devices is critical for risk mitigation. The Food and Drug Administration's (FDA) Manufacturer and User Facility Device Experience (MAUDE) database serves as a key tool for post-market surveillance, receiving reports from various sources. Ensuring information integrity, especially across diverse reporting sources, is paramount. Periodic regulatory updates aim to enhance reporting.

Objective

Analyze reporting patterns, assess the potential contribution of 2015 and 2018 regulatory updates on reporting rates for varies reporters. Additionally, evaluating reporting schedule compliance by comparing average reporting times to regulatory requirements for these reporters.

Methods

Data from 1991–2022 was retrieved from the MAUDE database. Annual report counts were filtered by reporter and event type. Average reporting time was calculated as the difference between received and awareness dates.

Results

The 2018 Voluntary Malfunction Summary Reporting (VMSR) program correlated with an increase in manufacturers’ event reporting, peaking at 3,135,501 events in 2022. Distributors’ reports surged post-2018, while voluntary reports steadily increased since 1997, spiking notably in 2015 with the Electronic Medical Device Reporting (eMDR) submission update. Reporting times for deaths averaged 80 days for manufacturers, far exceeding regulatory requirements, compared to 40 days for distributors and 46 days for user facilities. Malfunction events had longer average reporting times for manufacturers (89 days) compared to distributors (44 days).

Conclusions

Changes in reporting trends around 2015 and 2018 may be linked to regulatory updates. Tailoring regulatory changes for each reporter type, like user-friendly electronic forms, new reporting programs, and exemptions cancelations, can improve reporting. Balancing the volume of reports among different reporters is crucial for enhancing database integrity, transparency, and overall health outcomes.

Introduction

Medical devices play a crucial role in saving lives and aiding in the treatment of various conditions, leading to an increasing utilization of these devices over the years. However, the occurrence of adverse events remains a significant public health concern. Post-market surveillance involves the collection and analysis of adverse event data in dedicated databases, enabling regulators to identify safety signals and initiate actions such as recalls and warnings as necessary. In the US, the MAUDE database serves as a vital resource for the FDA, providing critical signals and decision-making basis for regulatory actions.^1,2

Recent research emphasizes that the underlying data received from the database need to meet high-quality standards to ensure a timely safety signal generation. ¹ Data integrity can be impacted by several factors, with under-reporting being a major concern that results in incomplete information. ³ Under-reporting is a pervasive issue affecting post-market surveillance activities globally.^4,5 Previous studies have identified multiple factors contributing to under-reporting, including limited awareness of reporting systems, fear of consequences, uncertainty regarding reporting requirements, and time constraints.^6–8 Data integrity can also be compromised by incomplete reporting forms submitted by reporters. ⁹ Efforts to address under-reporting and data completeness in the MAUDE database, include expanding the FDA's MedWatch program, which facilitates reporting by health professionals, patients and consumers. ¹⁰ Additionally, the FDA provides resources to assist reporters, such as the eSubmitter Tool to help ensure complete submissions and the CDRH Learn platform, which offers training modules on proper reporting procedures. Electronic Medical Device Reporting (eMDR) requirements mandate that reporters submit an updated report with any missing information as soon as it becomes available, ensuring that incomplete fields in the database are promptly updated and supporting the FDA's efforts to maintain data integrity. The FDA also provides detailed Medical Device Reporting (MDR) guidance and supports reporters through the eMDR system. Public awareness campaigns, industry collaborations, and data quality reviews also help to enhance reporting practices.

Diversity of data source can also influence data integrity. The MAUDE database incorporates reports from different reporters, including manufacturers, distributors, user facilities, and voluntary users. This diverse array of contributors can introduce variations in the data due to differences in device types, product problem indications, and device problem codes being reported. ¹¹ Nearly 97% of the reports in the MAUDE originate from manufacturers, ¹² a concentration shaped in part by the FDA's regulatory requirements ¹³ and reflecting a potential imbalance in data representation. ¹² While manufacturer reports incorporate input from a variety of occupations, such as physicians, nurses, attorneys and others, they ultimately reflect the manufacturer's perspective. In contrast, direct reports to the FDA from physicians, patients, other voluntary reporters, or health facilities provide unique perspectives that can offer a different view of the same events, as well as report events that may not come to the manufacturer's awareness. Enhancing reporting levels across these diverse sources is thus crucial to achieving a more balanced and comprehensive representation of adverse events within the MAUDE database.

Reporting levels and trends are susceptible to numerous factors, including recalls, ¹⁴ market expansion, clinical procedure frequency, and global events like the covid-19 pandemic, ¹⁵ each potentially impacting reporting entities differently. Studies, mainly in pharmacovigilance, suggest that legislative updates can influence reporting behavior, boosting volume, quality and transparency.^16,17 For example, the introduction of a user-friendly reporting form in the US in 2013 was associated with increased consumer reporting, ⁹ while new EU pharmacovigilance legislation adopted in 2010 appeared to empower consumer reports. ¹⁸ Nevertheless, the impact of regulatory updates on reporting patterns, especially from diverse sources, remains under-researched in the context of MAUDE. This paper aims to provide insights into these trends.

Our focus is on two recent updates to medical device regulations implemented in the US in 2015 and 2018 (as outlined in Table 1). The first update in 2015, eMDR, mandates that manufacturers and importers submit MDRs electronically to efficient processing, review, and archival. The eMDR system includes mandatory form fields designed to enhance data completeness. The second update, The 2018 Voluntary Malfunction Summary Reporting (VMSR) program, applies to manufacturers, allowing them to submit certain device malfunction MDRs in summary form on a quarterly basis, ¹⁹ with each report summarizing the total number of malfunctions. The summary reports are available to the public in MAUDE, intending to provide benefits for both the FDA and the public. This includes increasing transparency for the public by revealing events that were previously undisclosed,^20,21 helping the FDA to process malfunctions efficiently, enabling easy identification of malfunction trends, and reducing the reporting burden on manufacturers. Prior to the VMSR program, manufacturers were permitted to submit quarterly summaries of adverse events through the Alternative Summary Reporting (ASR) program; however, these summary reports were not disclosed publicly, and were maintained separately from the MAUDE database. ²² In 2019, the FDA revoked all such exemptions, and recently published the data on the FDA website to improve transparency, but the list of qualifying devices remained undisclosed. ²³ In summary, the regulatory updates of 2015 and 2018 introduced significant changes, particularly for manufacturers and distributors.

Table 1.

Summary of key regulatory updates in U.S. medical device adverse event reporting.

Update date	Effective date	Regulation update	Description of regulatory changes
1997		ASR (Alternative Summary Reporting)	Allowed manufacturers to submit summary reports for common and well-understood device issues instead of individual reports. These reports were not publicly available in the MAUDE database until ASR was eliminated in 2019. Manufacturers had to submit a special request to report in this manner. No specific device list was required for ASR reporting.
February 11, 2014	August 14, 2015	Medical Device Reporting: Electronic Submission Requirements (eMDR)	Introduced a user-friendly form with mandatory fields to enhance data accuracy; required electronic submission of reports by manufacturers and importers, discontinuing manual reporting.
August 17, 2018	Immediate	Voluntary Malfunction Summary Reporting Program for Manufacturers (VMSR)	Introduced a voluntary program allowing manufacturers to report certain malfunctions related to devices with specific product codes in a summary format on a quarterly basis, subject to specific conditions.
June 2019		ASR (Alternative Summary Reporting)	FDA removed ASR to improve transparency, making all reports on device issues publicly accessible in the MAUDE database.

A comprehensive understanding of reporting behavior is crucial, as it can foster increased reporting when supported by appropriate legislation. Existing literature tends to focus on individual devices or device categories. However, there is a need for a broader analysis of MDRs that considers the full spectrum of information reported in the MAUDE database. A recent study addresses this gap by examining sex-based differences and patient outcomes in adverse event reports. ²⁴ Our study aims to expand on this by conducting a comprehensive assessment of the reports in the MAUDE database, exploring differences among reporters. In this study, we aim to provide insights into reporting patterns in MAUDE both before and after implementation of new legislations, with regard to manufacturers, distributors, user facilities, and voluntary users. Specifically, we aim to observe and assess the potential impact of the two aforementioned legislative updates on reporting. In addition, we also evaluate the compliance of mandatory reporters with regulatory requirements. We specifically scrutinize reporting times as an additional informative factor. The reporting interval is defined as the duration from awareness of an event to actual reporting time. Legal obligations stipulate distinct reporting timelines for various reporters. By delving into these topics, we aim to enhance understanding of the reporting characteristics. This knowledge could help improve regulatory updates aimed at enhancing reporting and transparency.

Methods

A retrospective, observational analysis was conducted using data retrieved from the MAUDE database. The study spans a 31-year period from 1991 to 2022, with a specific focus on the years 2005 to 2022 and the impact of regulatory updates introduced in 2015 and 2018. Additionally, the study evaluates compliance with reporting schedules during this period. Reporting data was collected from the MAUDE database, which is publicly available on the FDA website. MAUDE comprises diverse datasets, each encompassing distinct details on reported events, which can be interconnected through a distinctive “MDR Report Key”. Python was employed for data organization and analysis. The data was retrieved from the mdrfoithru2022.txt file (CSV format), where each row represents a single report, with 82 columns detailing its properties.

In stage 1, we examined overall reporting trends across multiple years, and subsequently refined the analysis by categorizing the data based on different reporter sources (manufacturers, distributors, user facilities and voluntary users) using the REPORT_SOURCE_CODE column. The number of reports and summary reports submitted were counted annually, as well as the number of events reported in each summary report. Reports were included if they contained complete data (a total of 82 columns) and, for summary reports, if they had a value greater than 1 in the NOE_SUMMARIZED column, indicating that the submission reports multiple events. Reports with incomplete data (fewer than 82 columns) were excluded from the analysis. After applying the criteria, the analysis included 15,775,718 reports from 1991 to 2022. Removing reports before 2005 reduced the dataset to 15,182,755.

In stage 2, events were filtered based on reporter source (REPORT_SOURCE_CODE column) and event type. Event types were determined using the EVENT_TYPE column, which categorizes each event as a malfunction, injury, or death, according to the selection made by the reporter. Then, reporting time was computed by measuring the time difference between the date the event was received by the FDA (RECEIVED_DATE column) and the date the reporter first became aware of the occurrence (DATE_MANUFACTURER_RECEIVED or DATE_FACILITY_AWARE columns). Subsequently, the average reporting time and standard errors for each reporter were computed for 2005–2022.

Inclusion criteria included all reports containing complete data (a total of 82 columns) and those with a value in the REPORT_SOURCE_CODE column (U = user facility, D = distributor, or M = manufacturer). Initial reports, where the value in the TYPE_OF_REPORT column is I, were included, as well as reports with valid date values in the column: RECEIVED_DATE, DATE_MANUFACTURER_RECEIVED, and DATE_FACILITY_AWARE, which indicate when the FDA, manufacturer, and user facility or distributor, respectively, became aware of the event. Exclusion criteria encompassed reports with incomplete or missing data, updated or final reports where the value in the TYPE_OF_REPORT column is not equal to I, as these may have altered the received date, leading to inaccurate or negative date differences. Reports where the date differences were negative, indicating incorrect data, and reports with a year earlier than 2005, were also excluded. After applying the criteria, the analysis included 8,784,969 reports for manufacturers, 288,477 for distributors, and 9392 for user facilities. Further details on the methodology are available in Appendix 1.

Results

In the period 1991–2022 a prevailing upward trend is evident in submissions to MAUDE from diverse sources as depicted in Figure 1. However, this trajectory is marked by varying growth rates across the years. Concerning the 2015 and 2018 regulatory updates, the total number of reports is delineated across three distinct periods. There is a moderate upward slope in reporting between 2005 and 2014, preceding the 2015 update. Subsequently, a more pronounced incline is evident between 2016 and 2021, post the 2015 update, and this trend further accelerates after the 2018 update, indicating a heightened reporting rate during this interval. Notably, between 2014 and 2016, a deceleration is observed, suggesting an adaptation period following the 2015 update. As mentioned, following the 2018 update, a single report in MAUDE could encompass more than one event (periodic summary report). Light gray dots signify total reports submitted to MAUDE, encompassing periodic summary or regular reports, whereas dark gray dots represent the total reported events, inclusive of those reported within a single submission of summary reports. Notably, in reports after the 2015 update until the 2018 update, there is evidence of instances where a single submission represents more than one event, and the number of events seems to have increased following this regulatory update.

Figure 1.

Trends in the number of reports^a and events^b in the MAUDE database (1991–2022)^c. Arrows indicate the years when legislative updates occurred (2015, 2018). ^a Total number of reports submitted to the MAUDE database each year. This includes both individual reports and summary reports. Each summary report is counted as a single report. ^b Total number of events reported. An event is any undesirable experience related to the use of a medical device, potentially resulting in death, injury, or malfunction, as determined by the FDA. In the MAUDE database, most reports contain one event, but some reports, including summary reports, may include multiple events, which explains the difference between the number of reports and events. ^c The data, spanning the years 1991–2022, were extracted from the MAUDE database and include all types of reporters.

Figure 1 reveals significant increase of reports influx over the years, reaching a peak of 3,135,501 events reported in 2022. This total is derived from 2,945,026 submissions to MAUDE in that specific year. Consequently, we thoroughly scrutinized each reporter's contribution to the volume of reports, starting from 2005 when reporting became more extensive. This analysis aimed to elucidate the proportional representation of various reporters within the database. In the period 2005–2022, manufacturers submit the majority of reports (96.5%) to the MAUDE database, distributors account for 2.2% of reports, voluntary users account for 0.7%, and user facilities for only 0.5% (Table 2) . Hence, it is crucial to assess the conduct of each reporter individually and evaluate his respective contributions. While focusing on each reporter separately (Figure 2), there is a general upward trend over the years among all reporters regarding the number of reports, while their patterns vary significantly. Manufacturers consistently mirror the overall upward trend observed in Figure 1, constituting approximately 97% of the reports in the database. Distributors consistently maintained a low report count over the years (100s-1000 s); however, following 2018, their submissions experienced a dramatic surge after the new VMSR program introduced. User facilities and volunteer reports show fluctuating patterns with both increases and decreases over time. The number of voluntary reports has consistently increased since 1997, with a notable acceleration in 2015 while the eMDR submission was entered. However, starting in 2016, their reporting volume shows a less consistent trend.

Table 2.

Distribution of reports and event types by reporter type (2005–2022) in the MAUDE database.

Reporter	Total reports	Death	Injury	Malfunction	other
Manufacturer	14,656,707 (96.5%)	159,689 (1.1%)	4,827,400 (32.9%)	9,571,234 (65.3%)	98342 (0.7%)
Distributor	332,261 (2.2%)	788 (0.2%)	314,860 (94.8%)	14,031 (4.2%)	2581 (0.8%)
Voluntary	110,369 (0.7%)	2041 (1.9%)	63,036 (57.3%)	37,352 (33.9%)	7643 (6.9%)
User facility	83,418 (0.5%)	2166 (2.6%)	11,143 (13.4%)	63,981 (76.7%)	6126 (7.3%)

Figure 2.

Yearly report submissions by reporter type (1991–2022): (a) manufacturers, (b) distributor, (c) user facilities, and (d) voluntary reports. The arrows indicate the years of legislation updates (2015, 2018).

Regarding the 2018 legislation update, in the period 2015–2022, manufacturers submitted summary reports (Table 3). An upward trend in the number of summary reports can be observed until 2019, followed by a slight decrease in 2020. However, there was a subsequent recovery in 2021, reaching a peak of 3239 summary reports in 2022. Summary reports predominantly consist of injury or malfunction incidents. A notable upward trend in reported events can be seen until 2019 with 119,375 events, followed by a significant decline in 2020 and a subsequent recovery in 2021 and 193,592 events in 2022. Additionally, an increase in their volume as a percentage is seen over the years until 2019, decrease in 2020 following subsequent recovery. The leading manufacturers that reported in summary report manner since 2016 are Stryker (5216 summary reports), Animas Corporation (762), Edwards Lifesciences (462), Baxter (368) and Medtronic (318). In addition, numerical typing errors in the form of “123” were identified in the database concerning the count of events presented in the summary reports.

Table 3.

Summary reports and events ^a data submitted to MAUDE (2015–2022).

Year	2015	2016	2017	2018	2019	2020	2021	2022
Number of summary reports submitted by manufacturers	1	585	867	1505	1972	1701	2192	3239
Number of events disclosed in manufacturers’ summary reports	123	12,041	46,619	90,569	119,375	35,311	75,563	193,592
Percentage of events in summary reports	0	1.4	4.8	8.1	8.5	2.3	3.8	6.5

^aTotal number of events reported within summary reports. Each summary report contains multiple events.

We further explored another regulatory aspect concerning adherence to the FDA's reporting requirements, with a specific emphasis on the compliance with reporting schedules across different reporters. The analysis was limited to reporters mandated to submit reports within specified timeframes. In the period 2005–2022, the average reporting time, from the date of awareness to the reporting date, varied among different reporters, as illustrated in Figure 3. The longest reporting times are observed among manufacturers, who are required to report within 30 days in all cases. Manufacturers demonstrate a notable disparity in the reporting timelines for deaths, with an average reporting period of 80 days and for malfunctions-89 days. They report injury cases within an average of 58 days. Distributors, who are required to report death or injury events within 30 days, have an average reporting time of 40 days for death events, while injury events are reported on average within 16 days. However, health facilities, who are mandated to report within 10 days, have longer reporting times ranging from 46 to 51 days for death or injury events. Standard errors appear on the graph. No trends in reporting times for manufacturer, distributor, or user facility reports were observed before or after the implementation of the legislative updates.

Figure 3.

Average reporting times by reporter type (a) and event type (b) compared to legal requirements (c). (a) Types of reporters: manufacturers, distributors, and user facilities. (b) Types of events: death, injury, malfunction. (c) The data, extracted from the MAUDE database for the years 2005–2022, were averaged for each reporting entity. The average reporting times for events categorized as death, injury, and malfunction were compared to the corresponding regulatory reporting time requirements.

Discussion

Our study revealed that reporting patterns among various reporters over the years are closely aligned with legislative updates. Notably, in 2015, the introduction of eMDR, ²⁵ and in 2018, the implementation of VMSR program. ¹⁹ We conducted a thorough analysis of overall reporting behavior, subsequently shifting our focus to each individual reporter considering their respective contribution to the overall volume of reports, emphasized the substantial volume of reports from manufacturers in the database compared to reports from other sources. This approach allowed us to gain insights into how different reporters responded to regulatory updates, and examine the potential impact of both legislative updates. Additionally, we identified key industry players who embraced the summary reporting feature. Furthermore, we scrutinized the average reporting time of each reporter and assessed their compliance with regulatory reporting requirements. In the following paragraphs, we will delve into a detailed discussion of these results.

The contribution of regulatory updates to the total reporting

Many factors can contribute to the total volume of reports FDA receives, such as amount of device on the market, device associated recalls, amount of medical procedures, legislation, under-reporting and pandemics. ¹⁶ In this study, we focused on FDA legislation updates and their potential contribution. The implementation of eMDR in 2015 likely played a role in the accumulation of reports within the database. The VMSR Program in 2018 also contributed to the event volume of manufacturers by reporting in summary reports. The manufacturers, who make up a significant share of the reports in the database (96.5%), mainly represent the general reporting behavior. ¹² The steeper upward trend observed between 2016 and 2022 can be attributed to the legislations implemented in 2015 and 2018, which significantly bolstered the volume of reports submitted by manufacturers. Although the data is not normalized, this assumption seems plausible. The transition to eMDR, following an initial adjustment period, resulted in an enhanced reporting level. Summary reporting, a feature enabling manufacturers to consolidate multiple events with the same device into a single report, existed in the ASR program before, but it functioned as an exemption and remained undisclosed in the MAUDE database, going unnoticed. ²⁰ Hence, with the adoption of the VMSR, there is a definite expectation of an upsurge in the number of published events in the database, accompanied by an enhanced level of transparency for the public. This aligns with findings from studies conducted in the pharmacovigilance field following regulatory updates.^9,18 Despite the increase in the total number of reports during the years following the legislation updates, Table 3 shows a temporary decrease in the summary reports and events in 2020. This could be related to the outbreak of the covid-19 epidemic, which triggered significant changes in daily behavior within personal and organizational frameworks in health care facilities around the globe. ²⁶ Health frameworks and institutions dealt with substantial and unexpected loads, which led the system to defer treatments of chronic morbidity, diagnosis and inquiry processes, elective treatments, preventive medicine and more.^27,28 The devices reported in the summary reports often represent the procedures described above, the scope of which has decreased during the covid-19 period. Previous work that examined reporting behavior of adverse events during the covid-19 supports the findings of this study. ¹⁵ Identifying numerical typing errors in summary reports underscores the importance for manufacturers to prioritize the precise entry of verified event quantities rather than inputting arbitrary numbers.

Reports volume

Most of the reports in the MAUDE database are from manufacturers and this in itself can create an unbalanced information. ¹² The differences in reporting rates may be attributed, in part, to the FDA reporting requirements. Manufacturers are required to report all adverse events, including malfunctions, injuries, and deaths, within 30 days of becoming aware of the event. Distributors are required to report deaths and injuries to the FDA within 30 days, but they are not required to report malfunctions. User facilities must report death events within 10 days, while for injury events, they are required to report to the manufacturer, with reporting to the FDA only required if the manufacturer is unknown. ¹³ User facilities are not required to report malfunctions to the FDA, but evidence shows they have. Yet, the user facility and distributor reports accounted for only 0.5% and 2.2% of the total MAUDE reports respectively. However, voluntary users, who have the potential to significantly enhance data integrity, ²⁹ contribute only 0.7%. It raises questions of the current regulation's effectiveness for non-manufacturer reporting, and the homogeneity of the data. ¹² The regulatory updates, such as mandatory electronic forms, with required fields, improved both under-reporting and data integrity by ensuring more complete submissions. We observed increased reporting from both voluntary reporters and manufacturers after implementing the user-friendly electronic form. The VMSR program also addressed some under-reporting issues previously embedded in the database. However, reporting from user facilities and voluntary reporters remains lower compared to manufacturers. Under-reporting and incomplete data significantly affect the overall data quality of the MAUDE database, creating gaps that hinder accurate assessments of device safety. This discrepancy points to a persistent integrity issue arising from varying levels of reporting among different sources, which warrants further attention and improvement. Such discrepancies in reporting by various entities can lead to undetected device issues due to under-reporting with potential public health implications. An example of this is the case of Poly Implant Prostheses breast implants, which were ultimately removed from use and became a global public health crisis and regulatory failure. ³⁰ These differences highlight the importance of ensuring balanced and comprehensive reporting to safeguard public health. While the enforcement mechanisms of the MDR regulation include seizure, injunction, civil monetary penalties, and criminal prosecution under the authority of the Federal Food, Drug, and Cosmetic (FD&C) Act, addressing under-reporting may require additional steps. These could include stricter enforcement of reporting requirements, such as penalties for non-compliance, alongside enhanced training programs and better incentives for voluntary reporting. Expanding mandatory reporting scope and improving data verification processes could also help mitigate the issue. Raising awareness through hospital programs and public campaigns could increase participation, especially among user facilities and voluntary reporters.

Assessing the influence of regulatory updates on various reporters

The variance in reporting volume among the reporters prompted us to conduct individual assessments of each reporter's reporting behavior and to evaluate the regulatory contributions; undoubtedly, there is a substantial divergence in reporting behavior among the different reporters. Focusing on the 2015 legislation update, a profound contribution is particularly noticeable in the voluntary users reports (Figure 2(d)). After the update, the number of reports submitted by voluntary users experienced a significant increase. A similar results were found in studies done in the EU EudraVigilance data ¹⁸ and in the US food and drug administration adverse event reporting system, ⁹ where the reporting behavior of customers also rise following improvements to the reporting form, making it more user-friendly. However, the effect of the 2015 update was less pronounced on distributers and health facilities.

Distributors have a relatively constant level of 100–1000 s of reports over the years (1991–2018), and starting from 2019 the amount of reports increases significantly to 10,000s–100,000 s. Many factors can contribute to this sudden increase. Pertaining to program changes, from 1997 through June 2019, the ASR program was in effect. Since the program ended, the FDA revoked all exemptions and manufacturers must submit reports through MAUDE. This update appears to have influenced distributors. In previous findings, distributors were found to report extensively on dental implants (Authors et al. forthcoming), which were initially exempted from manufacturer reports until 2019. However, following the exemption's cancellation, these reports are now reflected in the database. In addition, other types of exemptions, that allowed two mandatory reporting entities (mostly manufacturer/distributor relationship) to submit information through MAUDE about an individual event in a single MDR that covered the requirements for each mandatory reporting entity, also ended in 2018. This change resulted in each mandatory reporting entity having to individually submit their reports through MAUDE, likely contributing to the increased volume of reports by distributors.

Relating to manufacturers behavior (Fig 2(a)), three distinct periods are noted, mirroring the overall trends in the database due to their substantial volume of reports; the slowdown between 2014–2016 can be related to an adaptation period following the 2015 legislation update. The revised reporting form provides a user-friendly digital interface; however, manufacturers may require an adjustment period compared to voluntary users who find it easy to report through MedWatch forms.

The 2018 VMSR program update, along with the 2019 cancellation of the ASR program, significantly changed how manufacturers and distributors report adverse events. Under the VMSR program, manufacturers were allowed to report multiple similar events in a summary format, rather than submitting individual reports for each event. Before 2018, although manufacturers technically had the option to report multiple events together, these reports were classified as exemptions and were not publicly accessible in the MAUDE database. The VMSR program formalized this process and made these summary reports available to the public. Consequently, from 2018 onwards, there was a noticeable increase in the number of events reported in the database. This change not only enhanced public access to information but also likely made the reporting process more efficient for manufacturers, leading to a higher volume of reported events.

Reporting time

Concerning adherence to reporting deadlines among diverse reporters, it is evident that distributers are most committed to the FDA reporting times requirements. Manufacturers’ reporting times, especially for malfunctions and deaths, are very far from the requirement. User facilities also do not meet the reporting times but are closer than the manufacturers. Death reporting times are notably lengthier for manufacturers and distributors compared to injuries. This extended duration may be attributed to the potentially more profound impact of fatal events on the device, necessitating more comprehensive investigations to determine whether the event is indeed related to the device and warrants reporting. Moreover, manufacturers demonstrate a significant allegiance to their devices, and the ramifications of device-related deaths are substantial. Consequently, this loyalty may result in delayed reporting of deaths, taking more time than necessary. In addition, legal actions following significant incidents can also impact the timeliness of reporting. Other study in pharmacovilance, compare between serious and non-serious cases, showed that time to report was longer in serious cases than in non-serious cases which support our results. ⁹ On the other hand, user facilities, perhaps because they lack any loyalty to the devices, show no notable differences in reporting times between injuries and deaths. Another potential explanation is that manufacturers do not receive event data directly since they are not the primary source; instead, the information usually originates in a health facility and is transmitted to the manufacturer after a certain period. Tadashi Toki et al. ⁹ clarify this can account for the extended reporting times compared to user facilities. Nevertheless, in this study, reporting times were assessed from the moment of awareness, not the event itself, and yet, manufacturers still exhibit longer reporting times from the day of awareness.

A potential explanation of the diversity in reporting times may stem from variations in the environments where reporters encounter events and decide to report them to the relevant authorities. ⁹ Yet, time to report is an interesting measure of promptness in reporting and may help to describe reporting behaviors. A recent study in the EU showed that median time to report an adverse drug reaction (ADR) was approximately 30 days in spontaneous reports, regardless of whether it was reported by patients or healthcare professionals. ¹⁷ Median time to report ADR by any reporter in the US was shorter than that in the EU. ⁹ These differences in time to report are likely to reflect differences in reporting pathways: while some reports in the EU arrived at the authority via companies or other regulatory agencies. Different reporting times may also be influenced by different cultures and histories surrounding the reporting systems and use of medical device, such as publicity around medical device, as a previous report suggested. ⁹ We assumed the regulatory update, especially the eMDR, might improve reporting times due to ease of use, despite unchanged timing requirements. However, no effect was observed. Effective enforcement measures may be needed to reduce reporting times.

Conclusions

Our analysis of adverse event reports in the MAUDE database has shown the characteristics of spontaneous reporting in the US. Reports exhibit varying trends among different reporters, encompassing differences in the share of reports from the database. The reasons for these variations depend on many and varied factors. Additionally, it appears that regulatory and program updates play a role in reporting rates, but it is noteworthy that each reporter's response to these updates varies. The 2018 VMSR update prompted manufacturers to extensively report events, increasing the number of events documented in the database and enhancing transparency for the public. Moreover, the 2018 updates resulted in a substantial increase in the volume of reports from distributors. Voluntary users positively responded to the 2015 regulatory update- the user-friendly reporting form, which likely bolstered their reporting capabilities. Additionally, it was observed that reporting times for death events, tend to be prolonged, particularly among manufacturers. This delay is likely attributed to their commitment to their devices.

Little prior research has been conducted in this area regarding the MAUDE database, making our research novel and significantly contributive to understanding report's composition of different reporters. The study does not definitively establish the role of MAUDE reporting but does add to the growing literature. These findings emphasize the need to address balance between the different reporters appropriately when utilizing spontaneous reports, particularly in the context of signal detection and decision-making for public health. The findings underscore the importance of promoting reporting among different reporters to achieve a balanced representation of all viewpoints, as well as the necessity of amplifying the voices of less prominent contributors, such as voluntary reports. This objective can be accomplished through initiatives such as transparency-enhancing programs and legislation, targeted campaigns tailored to the reporting preferences of each reporter, and the customization of forms to the specific needs of different reporters. By these, reporting process can became more user-friendly and the overall process can be improved, ensuring a more comprehensive and accurate representation of adverse events and by that improve public health.

Limitations

This study has several limitations. Because it relied on spontaneous reporting data, all the limitations inherent to spontaneous reporting are applicable. ⁹ Various factors can influence the quantity of adverse event reports and the patterns of reporting, including those that are difficult to assess solely based on our findings, such as prior information, environmental context, interests, and motivation. Additionally, the marketing data of medical device and the frequency of medical procedures performed annually can affect amount of reported adverse events. Had this essential information been available, data normalization could have been applied. However, due to its absence, data normalization was not feasible. Consequently, it becomes challenging to definitively attribute differences solely to regulatory updates or programs, as they may be influenced in part by other contributing confounding factors. ⁹ The rise in reports following regulatory updates necessitates a longitudinal examination. For instance, regulatory alerts have been extensively investigated as a potential confounder in numerous pharmacovigilance studies.^14,31,32 This consideration may apply similarly in the context of medical devices.