









































5 J Global Clinical Engineering Vol.4 Issue 2: 2021

Received December 26, 2020, accepted April 19, 2021, date of publication May 25, 2021

Compilation About Adverse Events Recorded in FDA/USA 
and ANVISA/Brazil Databases Through Models Available 
in the Literature Concerning Analysis and Prioritization of 
Actions for Medical Devices

By J. Carlos de Souza, S. Mehrpour, M. M. Ferreira, Y. L. Coelho, G. De Castro Vivas, D. D. Rodriguez, F. De Assis Santos. T. F. Bastos-Filho    

Federal University of Espírito Santo, Vitória, Espírito Santo, Brazil 

ABSTRACT

The development and application of medical technologies have grown steadily in all health fields, offering numerous benefits to 
users. However, adverse events, which may cause severe consequences for patients, also have increased. Technical and human 
factors that provoke dangers are related to the complexity of the devices, quality control in manufacturing, software, mainte-
nance procedures, materials, and mode of use. This work aims to present the main alerts, dangers, and failures and some ways 
to mitigate them related to the following medical devices: Defibrillators, Infusion Pumps, Physiological Monitors, Pulmonary 
Ventilators, and Ultrasonic Scalpels. For that, we performed an analysis of adverse events reported in the Food and Drug Ad-
ministration (FDA/USA) and the Brazilian Health Surveillance Agency (ANVISA) databases since 2016. Finally, we classified the 
events into different categories, according to their similarity. The results show a total of 3,100 cases registered in the FDA for the 
six types of medical devices addressed in this work and 75 cases registered in the ANVISA/Brazil for two of them. Based on the 
top ten health hazards provided by ECRI (2016-2020), this work contributes to understanding the most significant hazards of 
the previously mentioned devices and the main ways to mitigate these risks. Throughout our research, we found that the risks 
addressed in this work are common to several medical devices; therefore, there preventative measures to avoid them must be 
established, for example, training users to use and maintain the equipment, improving their quality, and also reporting adverse 
events to manufacturers. 

Keywords – Adverse Events, ANVISA, ECRI, FDA, Medical Devices, Manufacturing, Training, Maintenance.

Copyright © 2021. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY): Creative Commons - Attribu-
tion 4.0 International - CC BY 4.0. The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) 
are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is 
permitted which does not comply with these terms.

http://www.globalce.org
http://globalce.org
http://globalce.org


J Global Clinical Engineering Vol.4 Issue 2: 2021  6

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

INTRODUCTION

According to the World Health Organization (WHO), 
a medical device is an “apparatus, instrument, machine, 
software, material or another similar article, intended for a 
medical purpose” as monitor treatments, help people with 
disabilities, diagnose and treat illnesses¹. In the current 
COVID-19 pandemic, measures of prevention and control 
of health services have been defined by the Brazilian As-
sociation of Clinical Engineering, whose guidelines include 
checking the configuration and availability of Intensive 
Care Unit (ICU) beds and their primary devices: mechani-
cal ventilator, multi-parameter monitor, defibrillator, and 
infusion pumps, noting the need for staff training to use 
them. In addition, it is also necessary to identify defective 
or unused equipment due to a lack of parts or inadequate 
maintenance.² In this sense, clinical engineers play an 
essential role in managing fundamental medical devices 
for treating patients affected by the disease.³

Despite the importance and benefits of medical equip-
ment in health care, adverse events are also associated with 
them. In Brazil, the National Health Surveillance Agency 
(ANVISA) classifies adverse events like health problems 
caused to the patient by a device subject to a health 
surveillance regime, even used under recommendation 
from the manufacturer.4 These events can occur because 
the medical device environment is a complex system of 
human-machine interaction that requires understanding 
the environment and identify risk factors.5

Every year, the Food and Drug Administration (FDA/
USA) receives many thousand reports of suspected medi-
cal device-associated injuries, deaths, and malfunctions.6 
The FDA uses these reports to detect potentially related 
safety issues, monitor device performance, and contribute 
to benefit-risk assessments of these products.7 Since 1991, 
FDA has received more than 4.4 million adverse event 
reports.8 In addition, the ECRI/USA publishes the annual 
top ten of health hazards that assist in understanding 
risks in health procedures worldwide.

This work addresses risks associated with six pieces 
of equipment commonly used in ICUs. The first is the 
Automated External Defibrillator (AED, non-wearable), 
which uses external electrodes to analyze the patient’s 
electrocardiogram (ECG) and automatically deliver an 

electrical shock to treat ventricular fibrillation on victims 
of sudden cardiac arrest.9 The second, Direct-Current 
Defibrillator (low energy), delivers an electrical shock of 
up to 360J through paddles placed either directly across 
the heart or on the surface of the body, which is used for 
restoring normal heart rhythm in pediatric defibrillation 
or cardiac surgery.10,11 

The third piece of equipment is the Infusion Pump (IP), 
which perfuses medications or nutrients to the patient at 
a controlled amount; a health professional programs the 
rate and duration of fluid delivery using the equipment’s 
software.5,12 Fourth is the Physiological Monitor (PM), 
which is a device connected to the patient, able to identify 
clinical emergencies when vital signs like heart rate, blood 
pressure, and oxygenation exceed preset thresholds; in 
this case, alarms are activated.13 

The fifth is the Pulmonary Ventilator (PV), which in-
volves a breathing tube placed in the patient’s windpipe, 
connected to the mechanical ventilator, which delivers 
oxygenated air.14 PV is used during surgeries or treatment 
for lung disease, essential to treat respiratory failure 
caused by COVID-19. Sixth is the Ultrasonic Scalpel (US), 
which generates harmonic vibrations in a metal rod that 
denatures proteins, cuts tissues, and coagulates them 
simultaneously.15

Unfortunately, there are harms associated with the 
use of these medical devices. Estimates from 2008 to 
2017 have shown alarming results: defective medical 
devices may have caused more than 1.7 million injured 
patients and approximately 83000 deaths worldwide.16 

These data denote the importance of identifying types of 
failures, hazards, and their causes, as what can be done 
to reduce them. Thus, this work aims to present the main 
alerts, dangers, and failures related to the use of PV, IP, 
AED, DC-Defibrillator, PM, and the US. This overview of 
their main events can guide users for their most appro-
priate management and best practices when using these 
medical devices.

METHODS 

The ECRI’s top ten health technology hazards ranked 
annually (from 2016 to 2020) have guided our research 



7 J Global Clinical Engineering Vol.4 Issue 2: 2021

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

regarding the types of equipment that most offer danger 
to patients and the main ways to mitigate them. On the 
other hand, a review of documents published (since 2016) 
by FDA/USA and ANVISA/Brazil was also analyzed here, 
which report adverse events related to medical equipment. 
Several papers from the literature about that subject were 
analyzed too in our research.

The FDA database contains medical device reports sub-
mitted by mandatory reporters, manufacturers, importers, 
and facilities, in addition to voluntary reports by consum-
ers, health care professionals, and patients.7 The medical 
devices addressed in our research (AED - non-wearable 
and DC-Defibrillator - low energy; IP; Non-Continuous 
PV; PM - without arrhythmia detection or alarms; and the 
US) were searched in the FDA database within the period 
from January 1st, 2016 to April 30th, 2020. The cases from 
the FDA were classified into six categories defined in this 
study, which are shown in Table 3. It is worth mentioning 
that only reports of death and injuries for these devices 
were considered. The information found on the ANVISA 
databases is shown in Table 4. The searches for PV and 
AED were conducted considering the same period, and 
the alerts found were classified into three categories.

RESULTS

Main causes of failures in medical devices

In their historical development, medical devices have 
an increasing degree of complexity, with the development 
of new components and materials. This complexity im-
pacted the maintenance and performance of the devices 
and their reliability,17 which is directly related to the 
increased failure rate (Fig. 1).

The analysis of contributing factors in the appearance 
of faults demonstrates that causes are varied. Tables 1 
and 2 respectively show the classification of incidents 
according to studies by Amoore using ECRI database 
and Shepherd.18,19 In these tables, aspects as “device” 
are repeated, including manufacturing, materials, and 
maintenance. Another common factor is the “user” or 
“operator,” i.e., ignorance, inadequate technical training, 
and staff negligence.17

FIGURE 1. Reliability conditioned by technical complexity17 

TABLE 1. ECRI Classification of medical device incidents17

Device

1) Human factor design
2) Parts design unexpected failure
3) Deterioration failure that requires preventive 
maintenance (e.g., Battery)

Operator
1) Training and use error
2) Diverted attention
3) Criminal intent

Facility

1) Human factor design
2) Parts design; unexpected failure
3) Deterioration that requires preventive 
maintenance
4) Maintenance error

Patient 1) Active patient action affected the outcome
2) Patient’s condition affected the outcome

TABLE 2. Shepherd’s Classification of medical device 
incidents17

Device

1) Design error
2) Device or accessory failure
3) Improper maintenance / testing / modification
4) Manufacturing error

User

1) Device miss-assembly
2) Failure in pre-use inspections
3) Improper connection
4) Improper reliance on an automated feature
5) Incorrect clinical use and control settings

External 1) Electromagnetic or radiofrequency interference
2) Power Supplies (including gas)

Support 
System 
Failures

1) Error in hospital policy
2) Failure to train
3) Improper storage
4) Lack of competent accident investigation
5) Poor pre-purchase evaluation



J Global Clinical Engineering Vol.4 Issue 2: 2021  8

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

In the scientific literature, it is possible to identify mod-
els such as the Swiss cheese, proposed by Orlandella and 
Reason, which allow understanding the system failures, 
which arise when protection measures are overcome (Fig. 
2).20 In that model, the human aspect is highlighted, which 
occurs when the error originates from inadequate actions 
from health personnel due to fatigue, stress, inattention, 
and negligence. Regarding the system, it is possible to 
standardize the security measures taken from design, 
quality control, safety testing, maintenance throughout 
the life cycle, and adequate user training. Each aspect 
is equivalent to a cheese layer, representing barriers to 
errors and present fragilities. Therefore, it is crucial to 
scientifically determine which layers are involved in medi-
cal device failures and ensure that these “cheese holes” 
are not aligned, creating problems.17

Another model is Pareto analysis, which shows that 
many failures occur in critical devices, being possible 
to determine the causes, allowing focusing professional 
attention on the most relevant situations and corrective 
actions. This model showed that misuse, lack of mainte-
nance, and use by untrained personnel are the leading 
causes of medical equipment failures.17 

Both models contain promising elements, which were 
applied in our research, detailed in the sections that follow.

FDA Adverse Events

The data in Table 3 shows adverse events related to 
devices of general clinical use (with important application 
in ICUs), such as the equipment addressed in our research: 
PM, IP, AED, and PV. In addition, Table 3 also shows the 
US equipment used in surgical procedures. Several cases 
were reported in the USA, totaling 3,100 events between 
2016 and 2020. The highlights are the equipment AED and 
IP, which have 1,382 and more than 1,424 reports. The 

PV, US, and PM have, in that order, 187, 60, and 40 cases, 
respectively, whereas the DC-Defibrillator has only 7 cases.

The AED presented 831 cases of operating issues associ-
ated with malfunction and shock problems and problems 
in defibrillation and alarm errors. For this equipment, 77 
cases were related to monitoring problems with incor-
rect messages, and 58 cases of assembly or structural 
defects due to the defective connection and impedance 
problems. Cases of incorrect procedures were 59 due to 
inappropriate actions that lead to burns. Hazards were 
20 events of shock and burn to nurses and physicians. 
Finally, unknown reasons were 337 events. Regarding the 
DC-Defibrillator, which is activated manually, the seven 
cases were related to device operating issues generated 
by inappropriate shock. All events were related to severe 
cases, with four deaths and three injuries (Fig. 3).

For IP, we analyzed a total of 1,424 events related to 
injury and death. Most of them were related to device 
operating issues (913) due to stop working and failure to 
deliver medication. Still, flow obstruction and alarm error 
were also reported. The assembly or structural defects 
had 124 cases reported due to the component disconnec-
tion and broken devices. The monitoring problems, with 
30 cases, occurred due to incorrect messages on display. 
Unknown reasons were 306 events.

PV covers 187 events, with 73 being device operating 
issues that correspond to airway pressure and oxygen 
saturation defects. The 65 hazard cases were linked to 
loss of smell sense and respiratory distress. Assembly 
or structural defects were 43 cases related to broken 
pieces, connection of tubing problems, and inadequate 
humidification. The PM comprised 40 cases, ten device 
operating issues related to alarm problems, software, and 
electronic motherboard problems. Incorrect procedures 
were due to inadequate or insufficient training. Seven-
teen monitoring problems were due to inappropriate 
electrocardiograms and incorrect display messages. The 
US had 60 cases, 43 due to device operating issues linked 
to failure to cut, malfunction during surgery, and energy 
output problems. The remaining cases were divided into 
assembly or structural defects and hazards, with 10 and 7 
cases, respectively, including disconnecting components 
and fragmented material. 

FIGURE 2. The Swiss cheese model for events occurrence20 



9 J Global Clinical Engineering Vol.4 Issue 2: 2021

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

The FDA’s adverse events are again shown in Fig. 3, 
but in this case, separating death and injury provoked 
by the device. Again, the data are alarming for AED, with 
892 deaths and 490 injuries related to events. IP has 424 
deaths and more than 1,000 injury cases. Injuries were 
also more common than death for PV, PM, and the US. 

ANVISA Adverse Events 

The data from ANVISA/Brazil is restricted to national 
and international events with medical devices used in 
Brazil. The search on this public agency official page of-
fers gross values, often unrelated to the device. Thus, we 
identified 38 alerts for PV and 37 alerts for AED (Table 4).

PV presents most cases of device operating issues, 
totaling 26. Of these, 24 are related to display and oxygen 
sensor failure, incorrect ventilation, and stop working; 
and two cases were caused by problems with equipment 
alarm, such as sound-related problems. The assembly or 
structural defects were eight cases due to lack of solder-
ing on the plate, leading to loss of power and short circuit 
interrupting the ventilation. Manipulation or installation 
problems (four cases) occurred due to problems in the 
power panel of the ventilators. For this type of device, three 
alerts contained records of death and 11 cases of patient 
hypoxia, which could cause sequelae and lead to death. 

AED presented 29 cases of operational problems, 
such as electric shock error, cable failure, attenuated 
discharge in defibrillation, and alarm error, which could 
lead to death and injury of the patient. For assembly or 
structural defects, there were 8 cases of battery drainage 
and component failures.

TABLE 3. Adverse events reported in FDA databases during January 1st 2016 to April 30th 2020(1).

Medical Devices Assembly or 
structural defects

Device operating 
issues Hazards Incorrect 

procedures
Monitoring 

problems
Unknown 

reasons Total

AED 
(non-wearable) 58 (4%) 831 (60%) 20 (1%) 59 (5%) 77 (6%) 337 (24%) 1,382

DC-Defibrillator - 7 (100%) - - - 7
IP 124 (9%) 913 (64%) 10 (1%) 41 (3%) 30 (2%) 306 (21%) 1,424+ (1)

PM - 10 (25%) 10 (25%) 3 (7,5%) 17 (42,5%) - 40
PV 43 (23%) 73 (39%) 65 (35%) - 6 (3%) - 187
US 10 (17%) 43 (72%) 7 (11%) - - - 60

Total 235 1877 112 103 130 643 3,100

FIGURE 3. Graph of death and injury found in FDA referring 
to Table 3(1) 

(1) Considering up to 1,000 cases for IP related to injuries and all 424 cases related to death

TABLE 4. Adverse events reported in the ANVISA/Brazil data-
bases during January 1st 2016 to April 30th 2020.

Medical 
devices

Device 
operating 

issues

Assembly or 
structural 

defects

Manipulation 
or installation 

problems
Total

PV 26 (68%) 8 (21%) 4 (11%) 38
AED 29 (78%) 8 (22%) - 37
Total 55 16 4 75



J Global Clinical Engineering Vol.4 Issue 2: 2021  10

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

Top Ten Health Hazards of the ECRI

The ECRI is a nonprofit organization, which develops 
guidance for improving the safety and quality of care 
across all healthcare environments worldwide. Every year 
they produce a report of the top 10 health technology 
hazards, whose items represent hazards that managing 
technologies can minimize. ECRI’s engineers, scientists, 
and clinicians select topics based on insight gained through 
investigating, testing, observing operations, reviewing the 
literature, and speaking with clinicians, clinical engineers, 
administrators, and device suppliers.21-25

Comparing our results to ECRI lists, we noticed a con-
vergence regarding problems and errors presented by 
the six devices evaluated. Devices alarm problems were 
present in all five lists considered. For PV and IP, alarm 
malfunction, overload, and loss of alarms could induce 
severe consequences in patients. Infusion errors appear 
in the 2017 and 2019 lists. Problems with device opera-
tion by the medical team were listed in 2016 and 2019; 
however, many other cases were related to inadequate 
procedures.21-25 Regarding device cleaning, alerts were 
on the five lists due to patient infection or technical prob-
lems arising from incorrect cleaning. Structural problems 
appeared in 2020 (such as the risk of loose nuts and 
bolts to device failures) and 2019 (about device battery 
charging defects). From 2017 to 2020, cybersecurity risks 
were emphasized due to system exploitation by hackers, 
causing health care disruption.21-25

DISCUSSION

Nowadays, practically all health specialties need mod-
ern technologies, going beyond health establishments 
to patients’ homes. However, the risk of adverse events 
concerning these technologies is growing rapidly. These 
events can result from a single cause or the simultane-
ous occurrence of several factors, with the clinical team 
generally being held responsible. However, we identified 
several causes to be considered in all processes: the choice 
of technology, proper installation, technical maintenance 
throughout the life cycle, and correct use in relation to 
the patient. 

The results shown in Tables 3 and 4 suggest high reli-
ability. However, it is worth commenting that In this sense, 
Table 3 shows adverse events recorded at the FDA/USA 
whereas Table 4 presents alerts from ANVISA/Brazil.

The hazard for patients occurs when: an alarm condi-
tion is not detected by a medical device (such as IP, PM, 
or PV); the condition is detected but not communicated 
to a staff member, or the condition is communicated but 
not appropriately addressed.20 Regarding the PV, injuries 
occur mainly in the respiratory tract because the patient 
depends on this equipment for ventilation. Errors in the 
air supply, if not rectified, can lead to damage like hypoxic 
brain or lung injury and death, as shown in some records 
in this study. These devices have alarms that indicate in-
adequate ventilation, so proper configuration is needed. 
However, the challenge is to manage the alarms, which 
are usually missed due to alarm fatigue (when the team 
is overloaded), lack of sound sensitivity, or failure in the 
notification of alarms, in which they are not effectively 
communicated to staff.26

Other factors contributing to the inadequate ventilation 
implementation include insufficient knowledge of the best 
practices for ventilation and ventilator functionality.21 

Healthcare facilities need policies on setting ventilator 
alarms and protocols for verifying components. In addi-
tion, too often, lung-protective strategies and advanced 
ventilator tools are not commonly used, and best practices 
are not adopted.22,24 Mitigate these problems by verifying 
that all staff members dealing with mechanical ventilation 
have a good understanding of how these devices work.21

PM is used in physiological monitoring. The improper 
customization of the alarms could make it more difficult 
for the operator to understand changes in the patient’s 
physiological conditions or problems with the device. 
These systems must be configured not to act too many 
alarms or too few alarms, as this involves settings based 
on the needs of a care area and the patient’s condition. In 
addition, establishing policies and educating staff about 
optimal alarm-customization practices can help reduce 
the risks of loss sounds and harm to the patient.13,24,27



11 J Global Clinical Engineering Vol.4 Issue 2: 2021

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

A total of 424 deaths and up to 1,000 injuries (Fig. 
3) related to IP were recorded in this study. The incor-
rect programming procedure performed by the medical 
team occurs even with smart pumps that have a dose 
error reduction system. In this case, the patient can re-
ceive either too much or too little solution. The complex 
programming display and the absence of procedures to 
verify the programming can contribute to these errors. 
Thus, the surest way to eliminate them is to use auto and 
double-checks programming. Still, the staff needs to notice 
signs of damage to the IP components to guarantee the 
correct flow of medication.22,24

The AED has high values of death and injury, respec-
tively 892 and 490 cases in FDA. The relationship to the 
death of patients undergoing resuscitation is mainly linked 
to the operational failure of the device, for example, not 
providing an adequate charge or discharge. Successful 
defibrillation depends on delivering the shock to the 
myocardium, as the longer brain and heart are deprived 
of oxygen, the more damage suffers.10,11

The US is reported to be quick for the cutting and 
coagulation of tissue.28 Studies claim the benefits of this 
equipment, including allowing faster and safe surgical 
procedures.29 However, the alerts show that no device is 
exempt from technical and human failures; for example, 
there might be improper cutting.

Achieved results indicate the essential need for better 
protocols on activity verification and medical equipment 
quality control, especially for high-risk instruments. It is 
also necessary to provide medical staff training about the 
operation and execution parameters of all equipment to 
get good accuracy.11

Another critical point is the medical equipment 
maintenance carried out by clinical engineers. Thus, the 
predictive maintenance that accompanies equipment 
performance parameters, aiming to define the right mo-
ment of the intervention, with the maximum use of the 
asset, proves to be profitable, combining operational 
safety of the equipment and cost.30 On the other hand, 
preventive maintenance, according to NBR 5462-1994, 
“is carried out at predetermined intervals, or according 

to prescribed criteria, designed to reduce the probability 
of failure or degradation of the functioning of an item”31, 
therefore offering more safety.

In Brazil, to guarantee the safety and the values mea-
sured within the reliability standards of medical equipment 
and to obtain the Brazilian certification by the National 
Institute of Metrology, Quality and Technology (INMETRO), 
the clinical engineering team management performs 
testing and calibration of equipment following Brazilian 
standards, such as RDC number 02 and NBR15943.4,32  The 
manufacturers and distributors have a great responsibility 
in producing equipment in compliance with regulations 
and quality requirements. 

On the other hand, health authorities must follow 
regulations, conduct technological surveillance, and col-
lect information about events. In health establishments, 
the clinical and biomedical engineers are professionally 
trained to relate scientifically to devices, being increasingly 
important in product certification, choosing technologies 
and training of personnel, and thus helping to avoid seri-
ous failures.17

To evaluate the events addressed in our research, we 
used Pareto’s analysis to prioritize corrective actions 
and quantify the causes of problems in medical devices, 
allowing focusing the professional’s attention on the 
most relevant causes. The Swiss cheese model was used 
when protective measures of systems were overcome by 
circumstantial factors that combined them and produced 
an undesirable result. This model encompasses human 
aspects, such as faulty actions and the system, which 
need barriers against errors (cheese layers). The layers 
represent points in developing and using a device that can 
have weaknesses, so these layers cannot align.17

Finally, the alerts, hazards, and adverse events reg-
istered allowed us to identify the best practices to be 
adopted concerning the highlighted medical devices. This 
included increasing the training of operators and techni-
cians in maintenance, expanding predictive maintenance, 
changing the corrective maintenance modus operandi, 
adapting the infrastructure of the health care establish-
ment (hospital, clinic, polyclinic, etc.), identifying the need 



J Global Clinical Engineering Vol.4 Issue 2: 2021  12

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

to replace obsolete technologies, providing feedback to 
manufacturers and suppliers of medical technologies, 
and suggesting new public policies for the management 
of medical devices among other actions.

CONCLUSIONS 

The common faults in AED, DC-Defibrillator, IP, PM, PV, 
and US are related to alarm conditions not being issued 
by the medical device or not being adequately addressed 
by the team of professionals. In addition, these profes-
sionals are often not adequately trained to deal with the 
devices, the scarcity of system verification protocols, 
errors in the automatic execution of standard processes, 
lack of maintenance and programming according to the 
patient’s needs.

All medical devices can fail; however, the failures must 
be avoided by adequately selecting and maintaining these 
devices. For that purpose, it is necessary to pay attention 
to medical devices’ clinical and technical needs, perform 
regular equipment tests and maintenance, and medical 
team training. In addition, the medical devices must 
have adequate incorporation with an extensive search 
for suppliers, involving technical, clinical, budgetary, 
and infrastructure areas, allowing for a specification that 
meets the clinical, operational, and cost.

To understand and mitigate adverse events, this work 
shows that it is essential to apply models to analyze their 
causes, for example, Pareto’s analysis, which prioritizes 
corrective actions. In addition, it is necessary to stratify 
the types of adverse events for medical equipment, for 
example, using the layers of the Swiss cheese model to 
help understand which stages of development and use of 
the device contributed to the failures. 

Indeed, future research and studies with other inter-
national databases are necessary to widen the outcomes 
obtained in our research. Nevertheless, we believe that all 
aspects brought through applying models from Pareto’s 
analysis and Swiss cheese can impact the mitigation of 
these adverse events and, consequently, offer end-users 
safer medical devices and more effective health care.

ACKNOWLEDGMENT

The authors thank CAPES, CNPq, and FAPES for their 
scholarships, and UFES for technical support.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

REFERENCES

1. World Health Organization. Medical devices. WHO; 
2020. Available at: www.who.int/health-topics/
medical-devices#tab=tab_1

2. Nota Técnica Nº 01/2020. Orientações para enfrenta-
mento à pandemia de Covid-19. ABEClin, 2020. Available 
at: https://ced.ifmbe.org/images/COVID-19_ENGEN-
HARIA_CLI%CC%81NICA-Equipamentos_e_infraestrutura.
NOTA_TE%CC%81CNICA_N%C2%BA_01_ABECLIN_Ba-
hia_2.pdf

3. Freitas E V S, Artega M A P, Cardoso P A, et al. A Review 
About Equipment for Mechanical Ventilation in Inten-
sive Care to Combat COVID-19 and the Role of Clinical 
Engineers. In: XXVII Brazilian Congress in Biomedical 
Engineering, 2020, Vitoria. Proceedings CBEB2020, 
1054-1059. 

4. ANVISA. RDC nº 02: Gerenciamento de tecnologias em 
saúde em estabelecimentos de saúde, 2010.

5. Gao X, Wen Q, Duan X, et al. A hazard analysis of class 
I recalls of infusion pumps. JMIR Human Factors 
2019;6(2):e10366. doi 10.2196/10366

6. United States Food and Drug Administration. 
Data Mining at the Center for Devices and Radio-
logical Health. FDA; 2018. Available at: https://
www.fda.gov/science-research/data-mining/
data-mining-center-devices-and-radiological-health

7. United States Food and Drug Administration. Medi-
cal Device Reporting (MDR): How to Report Medical 
Device Problems. FDA, 2020. Available at: https://
www.fda.gov/medical-devices/medical-device-safety/
medical-device-reporting-mdr-how-report-medical-
device-problems.

http://www.who.int/health-topics/medical-devices#tab=tab_1
http://www.who.int/health-topics/medical-devices#tab=tab_1
https://ced.ifmbe.org/images/COVID-19_ENGENHARIA_CLI%CC%81NICA-Equipamentos_e_infraestrutura.NOTA_TE%CC%81CNICA_N%C2%BA_01_ABECLIN_Bahia_2.pdf
https://ced.ifmbe.org/images/COVID-19_ENGENHARIA_CLI%CC%81NICA-Equipamentos_e_infraestrutura.NOTA_TE%CC%81CNICA_N%C2%BA_01_ABECLIN_Bahia_2.pdf
https://ced.ifmbe.org/images/COVID-19_ENGENHARIA_CLI%CC%81NICA-Equipamentos_e_infraestrutura.NOTA_TE%CC%81CNICA_N%C2%BA_01_ABECLIN_Bahia_2.pdf
https://ced.ifmbe.org/images/COVID-19_ENGENHARIA_CLI%CC%81NICA-Equipamentos_e_infraestrutura.NOTA_TE%CC%81CNICA_N%C2%BA_01_ABECLIN_Bahia_2.pdf
https://www.fda.gov/science-research/data-mining/data-mining-center-devices-and-radiological-health
https://www.fda.gov/science-research/data-mining/data-mining-center-devices-and-radiological-health
https://www.fda.gov/science-research/data-mining/data-mining-center-devices-and-radiological-health
https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems
https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems
https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems
https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems


13 J Global Clinical Engineering Vol.4 Issue 2: 2021

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

8. United States Food and Drug Administration. OpenFDA. 
FDA, 2019. Available at: https://www.fda.gov/
science-research/health-informatics-fda/openfda.

9. Electronic Code of Federal Regulations. DC-defibrillator 
(including paddles)., 2021. Available at: https://www.
ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=3dfaca70be
1507ca8fdb5029de41c3d7&mc=true&r=SECTION&n
=se21.8.870_15300.

10. Electronic Code of Federal Regulations. Automated 
external defibrillator system., 2021. Available at: 
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SI
D=7d33d575fa607c396a47d06b8b5faa0b&mc=true
&r=SECTION&n=se21.8.870_15310.

11. Golpaygani AT, Movahedi MM, Reza M. A study on per-
formance and safety tests of defibrillator equipment. J 
Biomed Physics Engineering 2017;7(4):397.

12. Infusion Pumps. FDA, 2018. Available at: https://www.fda.
gov/medical-devices/general-hospital-devices-and-supplies/
infusion-pumps

13. Bonafide CP, Localio AR, Holmes JH, et al. Video 
analysis of factors associated with response time to 
physiologic monitor alarms in a children’s hospital. 
JAMA Pediatrics 2017;171(6):524-531. doi 10.1001/
jamapediatrics.2016.5123.

14. Pham JC, Williams TL, Sparnon EM, et al. Ventilator-
related adverse events: a taxonomy and findings from 3 
incident reporting systems. Respir Care 2016;61(5):621-
631. doi 10.4187/respcare.04151.

15. Crispi CP, Crispi Jr CP, da Silva Reis Jr, et al. Hemostasis 
with the ultrasonic scalpel. J Soc Laparoendosc Surg 
2018;22(4). doi 10.4293/JSLS.2018.00042.

16. Schabacker M. Awareness Alone Won’t Improve Medi-
cal Device Safety. ECRI; 2018. Available at: www.ecri.
org/press/improve-medical-device-safety.

17. Avendaño G. Critical importance of multilateral studies 
related with adverse events in medical devices. Health 
Technol 2016;6(3):213-227.

18. Amoore JN. A structured approach for investigating 
the causes of medical device adverse events. J Med 
Engineer 2014.

19. Shepherd M. A systems approach to medical device 
safety. In: Clinical Engineering Handbook, Ed. J. Dyro 
Pub: Elsevier Academic Press; 2004:246-249.

20. Reason JT. Human error: models and management. 
Bmj, 2000;320(7237):768-770.

21. ECRI. Top Ten Technology Hazards for 2016. Health 
Devices, 2015. Available at: https://www.ecri.org/
Resources/Whitepapers_and_reports/Top_Ten_Tech-
nology_Hazards_2015.pdf.

22. ECRI. Top Ten Technology Hazards for 2017. Avail-
able at: Health Devices, 2016. https://www.ecri.org/
Resources/Whitepapers_and_reports/2016_Top_10_
Hazards_Executive_Brief.pdf.

23. ECRI. Top Ten Technology Hazards for 2018. Health 
Devices, 2017. Available at: http://www.ecri.org/
Resources/Whitepapers_and_reports/Haz17.pdf.

24. ECRI. Top Ten Technology Hazards for 2019. Health 
Devices, 2018. Available at: https://www.draeger.com/
Library/Content/Haz_18.pdf.

25. ECRI. Top Ten Technology Hazards for 2020. Health 
Devices, 2019. Available at: https://www.ecri.org/
Resources/Whitepapers_and_reports/Haz_19.pdf.

26. Evans R S, Johnson K V, Flint V B, et al. Enhanced noti-
fication of critical ventilator events. J Am Med Inform 
Assoc 2005;12(6):589-595. doi 10.1197/jamia.M1863.

27. Ruppel H, Funk M, Whittemore R. Measurement of physi-
ological monitor alarm accuracy and clinical relevance 
in intensive care units. Am J Crit Care 2018;27(1):11-
21. doi 10.4037/ajcc2018385.

28. Peng M, Meng Z, Yang Z H, et al. The ultrasonic harmonic 
scalpel for circumcision: experimental evaluation us-
ing dogs. Asian J Androl 2013;15(1):93. doi 10.1038/
aja.2012.67.

29. Ferri E, Armato E, Spinato G, et al. Harmonic scalpel 
versus conventional haemostasis in neck dissection: 
a prospective randomized study. Internat J Surg Oncol 
2013. doi 10.1155/2013/369345.

30. Otani M, Machado W V. A proposta de desenvolvim-
ento de gestão da manutenção industrial na busca da 
excelência ou classe mundial. Revista Gestão Industrial 
2008;4(2):1-16.

https://www.fda.gov/science-research/health-informatics-fda/openfda
https://www.fda.gov/science-research/health-informatics-fda/openfda
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=3dfaca70be1507ca8fdb5029de41c3d7&mc=true&r=SECTION&n=se21.8.870_15300
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=3dfaca70be1507ca8fdb5029de41c3d7&mc=true&r=SECTION&n=se21.8.870_15300
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=3dfaca70be1507ca8fdb5029de41c3d7&mc=true&r=SECTION&n=se21.8.870_15300
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=3dfaca70be1507ca8fdb5029de41c3d7&mc=true&r=SECTION&n=se21.8.870_15300
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=7d33d575fa607c396a47d06b8b5faa0b&mc=true&r=SECTION&n=se21.8.870_15310
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=7d33d575fa607c396a47d06b8b5faa0b&mc=true&r=SECTION&n=se21.8.870_15310
https://www.ecfr.gov/cgi-bin/retrieveECFR?gp=&SID=7d33d575fa607c396a47d06b8b5faa0b&mc=true&r=SECTION&n=se21.8.870_15310
https://www.fda.gov/medical-devices/general-hospital-devices-and-supplies/infusion-pumps
https://www.fda.gov/medical-devices/general-hospital-devices-and-supplies/infusion-pumps
https://www.fda.gov/medical-devices/general-hospital-devices-and-supplies/infusion-pumps
http://www.ecri.org/press/improve-medical-device-safety
http://www.ecri.org/press/improve-medical-device-safety
https://www.ecri.org/Resources/Whitepapers_and_reports/Top_Ten_Technology_Hazards_2015.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/Top_Ten_Technology_Hazards_2015.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/Top_Ten_Technology_Hazards_2015.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/2016_Top_10_Hazards_Executive_Brief.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/2016_Top_10_Hazards_Executive_Brief.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/2016_Top_10_Hazards_Executive_Brief.pdf
http://www.ecri.org/Resources/Whitepapers_and_reports/Haz17.pdf
http://www.ecri.org/Resources/Whitepapers_and_reports/Haz17.pdf
https://www.draeger.com/Library/Content/Haz_18.pdf
https://www.draeger.com/Library/Content/Haz_18.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/Haz_19.pdf
https://www.ecri.org/Resources/Whitepapers_and_reports/Haz_19.pdf


J Global Clinical Engineering Vol.4 Issue 2: 2021  14

Carlos de Souza, Mehrpour, Ferreira, Coelho, De Castro Vivas, Rodriguez, De Assis Santos. Bastos-Filho: Compilation 
About Adverse Events Recorded in FDA/USA and ANVISA/Brazil Databases Through Models Available in the 

Literature Concerning Analysis and Prioritization of Actions for Medical Devices

31. ABNT, NBR 5462. Confiabilidade e Mantenabilidade; 
1994.

32. ABNT, NBR 15943. Diretrizes para um programa de 
gerenciamento de equipamentos de infraestrutura de 
serviços de saúde e de equipamentos para a saúde; 2011.


