Internet of Things as support to reduce hospital errors related to medication administration

Objective: The high potential of digital technologies to reduce medication errors in patients hospitalized or under hospital care was explored in this investigation. The study identifies common types of medication errors and their probable causes. Furthermore, examines the technological innovation solutions available that have not yet been applied in this context, specifically the Internet of Things usage to reduce medication errors through a practical example. Methods: A rapid review was applied, exploring the papers from two databases. The analysis was limited to the period from 2017 to 2021, resulting in 147 articles from the VHL Regional Portal – Virtual Health Library and 257 articles from the PubMed database. Results: There were analyzed 40 studies. An error mapping related to the theme was made. Also, the technology application and their effectiveness were identified in the studies. According to the research, inattention or distraction due to the excessive journey of work was pointed out as the main reason that leads to medication errors with the systems employed. Conclusion: There are several opportunities to improve hospital procedures with new technological approaches. By implementing innovative technologies, hospital medication errors can be managed more efficiently and reduce costs and waste with errors medication.


Introduction
Several factors may contribute to increasing the complexity of hospital operations, including the diversity of patients and their health conditions, as well as the high number of people assisting patients in medical procedures, leading to an increased proportion of operations being performed incorrectly.These issues can result in an aggravation of the health conditions of patients already debilitated by some disease (Kim et al., 2020).The scientific literature identified different records of hospital errors related to patients.Overall the years' several researchers in the medical and nursing area have tried to get to the true causes and propose solutions for their mitigation (Prado & Vilela, 2019).With regards to clinical biomedical engineering concepts and new technologies, this work contributes to the conversation about how new paradigms might be applied.In this context, the Internet of Things (IoT) can be used to improve health services, especially hospital care.Through the Internet of Things infrastructure, electronic devices, and distributed cloud systems can be accessed to enhance health assistance through communications resources and other technologies.(Firouzi et al., 2018).There are several opportunities to develop innovative solutions with the IoT.In the hospital setting, for example, the development of new tools and instruments to improve medical decision-making is just one example (Morales et al., 2021).Hence, medication errors can be reduced.Concerning patient care, several issues have been reported within the scope of organizational losses, such as day-to-day situations, and human restrictions that can eventually result in errors in the treatment of patients that can produce temporary or permanent negative effects on their health.As a highly complex system, health care involves several critical situations with potentially life-threatening consequences.Recently, in the COVID-19 pandemic, the multidisciplinary health professionals' teams were exposed to the infection risks and distress situation reported in some papers including nurse's suicides in Italy (Lupia et al., 2020).These features can lead to errors during patient care (Rosa & Perini, 2003).In addition, the overload of the health system suffered in recent years has aggravated the situation, considering the inadequate working conditions and stress sorrowed by health workers in this period (Salazar de Pablo et al., 2020).
Medications are essential to health care and have been used directly in the handling of patients.However, the pharmaceutical's inadequate management can cause significant adverse reactions, or simply be innocuous for the treatment, resulting in different conditions from those foreseen in the medical treatment prescript (Rabelo Néri et al., 2011).Medication errors in health care have been constantly discussed in the context of hospital care, reported negative consequences of short-term impact and result in high human damage which is very difficult to quantify, both for the patient and for the professionals and The inclusion and exclusion selection criteria were applied and quantified according to the PRISMA (https://prismastatement.org) diagram in Figure 1.Articles related to primary care were selected, considering works related to very specific areas and/or high complexity due to the research objective outside the research topic.There were selected 40 studies.

Answer to RQ1
According to an article published by Schwendimann et al (2018), based on data analysis from twenty-seven countries related to "Adverse Events", or EA's, in hospitals, it would have been possible to avoid 50% of the damage due to hospital errors (Schwendimann et al., 2018).In a more recent publication, adverse events are described as situations accompanied by a series of conditions that generate insecurity in patient care.However, studies related to the subject point to psychological and behavioral issues (M.C. Hsieh et al., 2021;Risør et al., 2017;Ting et al., 2020), leaving aside many issues related to management and organizational problems (Cabilan et al., 2017;Oliveros et al., 2017).Errors can occur during the medication process administration, such as stress by health professionals, and it represents an overload in the health system involving social and financial issues that affect everyone (M.-C.C. Hsieh et al., 2021).For example, coping with the COVID-19 pandemic in the world health system (Burlea-Schiopoiu & Ferhati, 2020).The investigated articles found distinct types of medication-related errors (Risør et al., 2017).The rapid review enables the organization of Table 2 highlighting the importance to identify the most recurrent drug administration errors in hospital environments as an outcome of this work.
Table 1 -Medication errors and classes found in the review that are mentioned in several references.Source: Authors.

Answer to RQ2
Table 3 summarizes the types of technologies reported in the studies.Some articles refer to mobile devices (smartphones) and apps.In many cases, experimental solutions with machine learning and fuzzy appear, and in some cases consider other types of technology as analytical solutions for data analysis.

Technologies
Related articles Barcode Medication Administration -BCMA Refs.(Barakat & Franklin, 2020;Burkoski et al., 2019;Cabilan et al., 2017;González-Bueno et al., 2021;Kennedy & Massey, 2019;Küng et al., 2021;Mulac et al., 2021;Risør et al., 2017;Van Der Veen et al., 2018) Electronic Health System -HER Refs.(Amato et al., 2017;Biltoft & Finneman, 2018;Debono et al., 2017;Griffon et al., 2017;Jurado et al., 2018;Kirkendall et al., 2020;Pontefract et al., 2018;Reinhardt et al., 2019;Shah et al.,  the cycles of the medication into the MBMS device first.After, the medication will be administered by the patient himself.The system sends an alarm to the patient's smartphone or emits an alarm at the time of taking the medicine.When a patient hears the alarm and approaches the device, the camera records the video based on the motion detected around the motion detection sensors and stores the behavior of the patient's medication.The system uses OpenPose, the human activity recognition part analyzes the patient's behavior in relation to recorded images and identifies the act of raising the arm to drink water.After detecting the action, the device discards the right amount of medication to easily provide the patient with medication.The weight of the drug is checked according to the doses taken by the patient.If the patient's behavior is compatible with the act of taking water and the weight of the drug converges to zero, it is considered that the drug was taken at the right time.The device sends weekly results to the doctor, allowing the monitoring of the patient's behavior and the patient's health status.The device uses the MQTT protocol to send the information.Furthermore, information to assist in disease management, such as foods that may or may not be ingested, is performed through a monitor connected to the device (Roh et al., 2021).The second article in the IoT-based system context is not exactly for medication administration but is a device that predicts patient safety.It is a wearable device for blood pressure monitoring, which analyzes in real time and predicts the occurrence of strokes through machine learning algorithms.The device can be integrated into a hospital system architecture allowing monitoring through a panel that displays the graphics of the device's monitoring and can be accessed through a web page (Wang et al., 2019).
The implementation of barcodes in the administration of medicines (Barcode Medication Administration -BCMA) associated with electronic registration systems for label of adverse medication events was explored by nine studies.The studies reported satisfactory results, but pointing out problems related to the use of technology as one of the notification of health teams.
Between the difficulties presented it was mentioned the information system manipulation during the daily activities.Another technology cited, which resulted satisfactory consequences, is the use of Electronic Health Records (EHR).In some cases, the use of electronic registration was associated with BCMA technology Refs.(Cabilan et al., 2017;Küng et al., 2021;Mulac et al., 2021;Risør et al., 2017).Health Systems, or Health Information System (HIS), are mentioned in two studies Refs.(Chien et al., 2021;Lu et al., 2017)

Discussion
During the investigation, two papers were encountered in the rapid review belonging to the IoT context.It is essential emphasizing IoT ecosystems allow the exchange of information through shared systems remotely, regardless of their distance and geographical location (Atzori et al., 2010).When the new paradigm is associated with the health area, several opportunities can be explored, and it has been possible to find different definitions of health involving this concept: Internet of Medical Things, Internet of Things Healthcare, Smart Healthcare, Healthcare 4.0, Smart Hospital, etc.In the literature, existing many terms related to the suitability of these new technologies in the health area.Furthermore, there are several new types of research explored in this theme, new sensors, biomarkers, devices, microprocessors, and new machine learning approaches specifically to healthcare (Al-Turjman et al., 2020;Habibzadeh et al., 2020;Qadri et al., 2020;Scarpato et al., 2017).In general, the IoT allows the combination of various areas of knowledge and applications, such as data acquisition, communication, data analysis, and uninterrupted connectivity.This allows objects (or things) to store, exchange, and combine data that results in useful information for human decision-making.Until this moment, there is no international standardization of use for the IoT-based system architecture applied to health (Morales et al., 2021).There is not even a consensus on the distribution of the number of layers needed for your deployment, but usually, they have very similar functions.There is a challenge to improve security and privacy issues, specifically in the health context.

Employing IoT to Reduce Medication Errors
An IoT-based system is proposed to help reduce medication errors.Three-layer IoT requires wrist wearables, smart medicine cases, and interactive displays that record events automatically and provide access to the database with respect to the patient's identity.Figure 3 illustrates an IoT-based system proposal.Source: Authors.
One of the essential points to avoiding medication errors is related to the identification system, for both patient and the nurse.There are several types of medication errors, including incorrect information about names, health conditions, allergies, and other information that the nursing team may need to evaluate the medication administration process.Among the attributes of an IoT-based system is the ability to gather data, so patients and health workers should wear wristbands with an electronic microprocessor, a small memory, and a wireless communication interface.In addition to being comfortable for the patient or care professional, the wristbands need to be adjustable, for instance, on the wrist, leg, or on the body in the appropriate place.
Various colors can be used to indicate standardized relevant information, such as allergies or chronic diseases.Nowadays, it is used today is a plasticized paper tape with printed identification and a barcode.This solution, although practical and inexpensive.
It does not allow the inclusion of other information and it needs a portable barcode reader to integrate with the computer system.
Furthermore, it is impossible to add or update the information.The barcodes increase the likelihood of errors occurring in the process.Electronic wristbands holding integrated electronics enable the updating of information automatically, including medication received, food, hygiene and other, i.e., the registration of care by nursing staff.The medication box may also include wireless communication to store information related to the drugs included in it.For example, it is possible showing information about the pharmacy sector, the name of the physician, the pharmacist, the nurse, and the patient, as well as to detail each of the medications along with the time and the way of handling and administration of medicine.In the pharmacy data load, the system automatically checks allergies or antagonistic conditions with the selected substances, and it will alert the pharmacist or the nurse.Most of the information is automatically captured, including the electronic signature of the pharmacy professional at the time of release for administration.For the drugs, RFID tags should be used to aid in automating the process.A hospital tablet in each bed, fixed to the wall or the hospital bed, detects the arrival of the professional and the box of medications and checks that they agree with the identification of the patient corresponding to that bed, issuing authorization or verification alerts.This element records the event and forwards it to the cloud to be part of the data collection and record.The approach would be employed by hospital management for other applications around costs and control of drug use.Other communication elements spread throughout the various sectors of the hospital would record the movement of patients and nurses during the 24 hours, seven days a week.As shown in the Table 3, IoT-based systems enhance medication administration security.
Table 1 -Rules that the IoT-based system could follow to ensure patient safety while administering drugs process.
Source: Authors.Inform and instruct the patient about the medications he receives Electronic systems are able to speak aloud and even send this information by electronic message, for example, to a patient's or family's smartphone.Check that the patient is not taking any medication sofa to the prescribed (clinical error)

Rules
It's a point to be analyzed.The viewfinder has video camera, which can be used.However, there are practical, ethical and cultural considerations to take into account.Investigate whether the patient suffers from allergies and rule out pharmacological interactions (clinical error) Known allergies are stored in the cloud.The color of the bracelets may evidence allergies, diabetes, and other conditions.However, unknown allergies should be evaluated by conventional means.
Wash hands before preparing and administering a medicine (procedure error) The presence of the nurse in hand washing can be detected, however it is difficult to address all situations, other than the issuance of an alert message on the display, remembering this standard and requesting confirmation that it was met.

Final Considerations
This paper presented a rapid review of medication errors with technology usage in hospital environments.The results obtained help to mapping a set of medical errors and information technologies in the care of patients in hospitals.This study is based on the development of the research proposal and directed the search for technologies that could help reduce adverse effects in hospitals, specifically medication errors.The potential of information technologies and specifically IoT-based systems have not yet found its place in hospital care, as observed in the results that point out only two works in this context.However, there is increasing interest in researching and evaluating the use of this resource to improve hospital procedures and processes.This technology may improve the hospital management process.The technology proposed in this article is feasible, available, and can be integrated as a pilot project in any hospital unit.
Finally, we emphasize some suggestions about future works.The scientific community needs to advance how to guarantee privacy for patients and how to include security mechanisms in the Internet of Things Healthcare (Ma et al., 2019;Qadri et al., 2020).There are rare important results about this theme and have been a vital investigation to increase the usage of new technologies in hospital management and safe patient (Ahmadi et al., 2019).Some techniques may be incorporated into the solution before deploying it for field testing, guaranteeing the efficient and correct use of new technologies.Another important issue is related to wearables and materials to employ in hospitals to manage the vital signs of the patient.There are significant approaches and resources to be explored associated with innovation and application of new technologies in healthcare and safe patients (Minh Dang et al., 2019;Prieto-Avalos et al., 2022).
, Ref. (Vélez-Díaz-Pallarés et al., 2017).Incorrect dosage per unit Incorrect administration period Identification of the patient in the medicine box Patient identification control Incorrect barcode reading Distraction of the health care team Incorrect use of drug automation systems Inadequate monitoring Error in the packaging of the medicine Incorrect route administration Psychological issues

Figure 2 -
Figure 2 -Interaction of devices in the nurse-patient system.
Patient Safety Correct medicine (clinical error) Check the dispensing in the pharmacy and administration in the room.Correct patient (clinical error) Automatic identification in the bed of medicines, room and patient, even in rooms of multiple beds.Correct route (procedure error) Indicated the procedure on the display.The nurse should tap the screen to confirm the route.Correct dosage (procedure error) Checked in the pharmacy and checked by the nurse in the room through the smart viewfinder.The nurse should tap the screen to confirm the dosages.Correct time (clinical error) Automatic event logs and prescription-related views.Generation of notifications about possible delays.Registration of administered medications (organizational administration error) Confirmation from the nurse.

Table 3 -
Types of technologies reported in the studies.