“To err is human” is the title of a major report made in 1999 on the quality of health care in the United States of America . This report contained disturbing statistics on the leading causes of death in American hospitals. Would they be correct?
In 2013 the statistics were updated and it was verified that the forecasts were always correct. Medical error is estimated to be the third leading cause of death in American hospitals after heart disease and cancer (Figure 1).
Figure 1. Illustration of the main causes of death reported in the report “ To err is human”, edited in 1999 and reviewed in 2013.
In the European context, the data show us that medical errors and adverse events occur in about 8% to 12% of hospitalizations . Spain, France, and Denmark have published incidence studies with similar data.
There is a scientific discipline capable of understanding the source of most human errors and difficulties. It’s called Human Factors.
It is the discipline that aims to improve the relationship between humans and other parts of a system. Initially, it focused on the design of cockpits, controls, and controls inside aircraft , but its applications are endless. It includes elements of psychology, anatomy, and physiology, engineering, design, and a host of other disciplines associated with human and machine sociotechnical systems (Figure 2).
Figure 2. Some of the components of the discipline of Human Factors: Psychology, social sciences, design, organizational management, engineering, physiology, anatomy. Source: Human Factors for Health and Social Care, White Paper (2018). Chartered Human Factors Specialists
The goal of Human Factors is to optimize the capabilities of human beings so that they achieve better performance, safety, health, and satisfaction when interacting with a system. For this, it is necessary to understand human capacities, their characteristics, and variability. Human Factors professionals analyze man-machine interactions to optimize processes, techniques, interactions, and communication.
In the hospital context, a good indicator of medical errors is reported incidents in official databases. These records are then analyzed to study the causes of patient safety incidents and also help define areas in which efforts to improve quality and human factors need to be concentrated.
With regard to medical device incidents, the EU now provides the EUDAMED database which, as the Food and Drug Administration (FDA) has done for several years, will make available the reports associated with certified and marketed medical devices.
Medical devices are prone to interaction errors for a variety of reasons. When reading some of these reports, one might think that “human error” or “user error” are the first causes of the incidents. However, the Human Factors approach takes a broader perspective of the problem and incidents and attempts to address all causes, such as poor product design. As part of efforts to move beyond error and guilt in patient safety incidents, the focus has fallen on the part that technology plays for the good or bad accomplishment of a task.
A change in technology changes the nature of the human task.
Case study – Infusion pumps
Software and interface problems are a major reason for system or device failures. In the hospital setting, the work of investigator Paolo Masci with infusion pumps  is an excellent illustration of how incidents are not just the result of human error.
Infusion pumps are devices used by doctors to inject fluids into patients, usually medicines or nutrients. In four years in the United States, the FDA received approximately 56,000 reports of adverse events related to infusion pumps. 500 of these incidents resulted in deaths.
Many of these adverse events were categorized as an error of use but were the result of problems in software design and the interface itself. As an error in use, we can consider errors in entering numbers, errors in the transcription of the prescription to the device, or failure to verify if the value was correctly entered.
To understand how technology is sometimes the basis of the human error, Paolo Masci has created a simulation tool that can be found by clicking in this link and must be accompanied in the following video made by the same researcher:
One notices that it is easy to make typos when the equipment itself does not have the same mental model as its users. Some equipment goes from 9 to 10, others from 9 to 0; some equipment has a decimal point, others do not; in some devices navigate with the side arrows, in others these arrows record or erase the last number. When, like health professionals, one deals with types of equipment of several manufacturers, each with its working model, in environments of high mental load, the errors that happen in the hospitals become more comprehensible.
Case study – Auditory alarms
Another interesting case study is in auditory alarms inside the hospital.
Medical device alarms account for 25% of all noise within the hospital.
There is an international standard that indicates how these alarms should be designed (IEC 60601-1-8: 2012). However, although recommended in this standard, these alarms have several problems. They are difficult to learn and are easily confused with each other.
These errors occur because the recommended sounds were based on a premise that would work at the outset: linking songs to problems. For example, Tony Bennett’s “I Left My Heart in San Francisco” rhythm could be used to signal heart problems. And so it came to pass. Although they consist of beeps, these beeps have different rhythms that, according to some researchers , would function as mnemonics for the events they represent (heart, medication, temperature, respiration, among others).
But this idea has never been tested with users before being standardized. The laboratory tests that were done subsequently revealed the same thing: these alarms are not good when used in real situations, with a lot of equipment with their own sounds, and with a high mental load from the clinical staff.
This standard will be updated by the end of 2019 and will feature new proposals for audible alarms for medical equipment. In this sense, the CCG is also improving the sound landscapes of hospitals, with the Paterson project that aims to use a user-centered methodology to design sound signals that are easier to detect, learn and locate.
The importance of standardization
Due to the increasing occurrences of adverse events caused by medical device interfaces, the US FDA has included human factors procedures and usability engineering in its medical device approval and certification processes.
The good news for the EU is that the new requirements of the Medical Devices Regulation will closely follow the requirements in the USA. This is important in the interests of global harmonization and the consistent increase in quality standards. IEC 62366 – Application of usability engineering to medical devices, will be part of the medical device certification process outside the US.
Health care will benefit from standardization and simplification of tasks, with protocols and checklists to reduce variability and the application of usability engineering in the development process. Finally, the entire technology implementation benefits from usability testing with real users.
The Computer Graphics Center has the domain of Perception, Interaction, and Usability responsible for the issues of Human Factors in the human-machine interaction. In addition to currently working on the design of new sound alarms for medical equipment, this group has already participated in the system certification process for the American market, has researchers with experience in the FDA error analysis process, and is currently a member of the MDevNet: Rede Nacional de Transferência de Conhecimento de Dispositivos Médicos network.
 Donaldson, M. S., Corrigan, J. M., & Kohn, L. T. (Eds.). (2000). To err is human: building a safer health system (Vol. 6). National Academies Press.
 Catchpole, K., & McCulloch, P. (2010). Human factors in critical care: Towards standardized integrated human centred systems of work. Current Opinion in Critical Care, 16(6), 618–622.
 Harrison, M. D., Masci, P., & Campos, J. C. (2018). Verification Templates for the Analysis of User Interface Software Design. IEEE Transactions on Software Engineering, 5589(c), 1–20. https://doi.org/10.1109/TSE.2018.2804939
 F. E. Block, “‘For if the trumpet give an uncertain sound, who shall prepare himself to the battle?’ (I Corinthians 14:8, KJV),” Anesthesia and Analgesia, vol. 106, no. 2, pp. 357–359, 2008.
About the author:
Usability Analyst of D.I.A. Perception, Interaction and Usability @CCG
Graduated in Experimental Psychology, she is currently a usability analyst in the field of applied research Perception, Interaction and Usability of the Center for Computer Graphics. She has been involved in product certification projects at the usability level and user-centered design; integrates the Human Factors team of automotive HMI, Health and ICT projects, having carried out validation studies of icons and gestures, among other methodologies; At the moment she is a PhD student in Ergonomics at the Faculty of Human Motricity (ErgoVR, FMH, University of Lisbon), where she investigates sound warning signs in shared work environments.