The pandemic has posed unprecedented challenges for patient care — difficulties faced each day by clinicians, clinical engineers, IT specialists, supply chain managers, and other healthcare professionals. As a result, ECRI notes in its 14th annual report on the top 10 health technology hazards, that a number of new threats to patient and staff safety have emerged.
In addition, the list from ECRI, a leading national patient safety organization, highlights the dangers posed by some developing technologies, such as artificial intelligence and 3D printing.
Here are the group’s Top 10 Health Technology Hazards for 2021.
Table of Contents
Complexity of Managing Medical Devices With COVID-19 Emergency Use Authorization
To meet urgent clinical needs during the pandemic, the Food and Drug Administration (FDA) has temporarily authorized the use of hundreds of medical devices. Under this authority, “FDA can designate previously unapproved products — or new uses for previously approved products — as acceptable for use during an emergency.”
However, ECRI notes, these devices may not be as safe or effective as devices approved through FDA’s normal clearance process. The reason is that the agency uses a lower standard for checking safety and effectiveness before granting an emergency use authorization (EUA).
“Healthcare facilities that use EUA devices face a complex challenge: They must manage inventories of EUA devices and their documentation, monitor each device’s status daily to determine whether the EUA remains active and unchanged, and determine what to do with these devices once the EUA ends.”
Fatal Medication Errors Can Result When Drug Entry Fields Populate After Only a Few Letters
To make drug search and selection easier, many medication ordering, storage, and delivery systems “allow the practitioner to enter only a few letters of a drug name before the system populates the drug selection field with a list of drugs to choose from.”
This drug searching and selection functionality exists in EHRs, CPOE systems, ambulatory prescribing systems, automated dispensing cabinets, inpatient and community pharmacy systems, and infusion pumps.
Designed to be a convenience, this feature can display similar-looking drug names as options, increasing the risk that users will mistakenly select an incorrect drug. In several cases, this has led to severe harm or the death of patients, ECRI says.
Rapid Adoption of Telehealth Technologies Can Leave Patients and Data at Risk
Because of the pandemic, there has been a rapid increase in the use of telehealth. However, as organizations transition to the new telehealth delivery models, “programs may struggle to provide sufficient user training, to coordinate patient care, or to overcome technology resource inequalities among patients.”
Failure to address these challenges could adversely affect patient care, lead to suboptimal treatment, increase risk of medical errors, or prevent certain populations from accessing care. Rushed implementation might also result in inadequate cybersecurity controls to protect health IT systems and patient data.
Imported N95-Style Masks May Fail to Protect Healthcare Workers From Infectious Respiratory Diseases
For healthcare workers exposed to aerosols from COVID-19 patients, an N95 respirator is vital equipment. However, some imported N95-style masks — especially KN95 masks imported from China — fail to provide the level of protection claimed. According to ECRI testing, more than 60% of imported N95 masks fail to block at least 95% of airborne particles.
ECRI suggests that healthcare facilities test N95 respirators not certified by the US federal government before using them in high-risk areas.
Relying on Consumer-Grade Products Can Lead to Inappropriate Healthcare Decisions
Consumer-grade pulse oximeters, blood pressure cuffs, glucose monitors, and other devices are being used not only in the home, but also in other healthcare settings, ECRI observes. “Patients and clinicians alike are increasingly interested in such products to provide some level of care when a traditional medical device is unavailable, inappropriate, inconvenient, or too expensive.”
During the pandemic, moreover, such devices have been used to reduce bedside visits and exposure risks, and to address medical device shortages.
Nevertheless, ECRI notes, these products shouldn’t be relied on to make healthcare decisions because their measurements may be inaccurate or misleading. “Most consumer-grade devices have not been through FDA’s medical device approval process,” the report says.
Hasty Deployment of UV Disinfection Devices Can Reduce Effectiveness and Increase Exposure Risks
Ultraviolet light can be used to disinfect surfaces and spaces to supplement normal cleaning and disinfection processes. But UV light must be used at the right wavelengths and for the right amount of time to inactivate microorganisms. Improper use of these devices not only renders them ineffective, but can also expose operators or bystanders to unsafe levels of UV radiation.
UV disinfection devices are not typically regulated by the FDA, because most are not considered medical devices. There is no standard protocol for demonstrating their safety and effectiveness, so healthcare organizations have to pay extra attention when purchasing these devices.
Vulnerabilities in Third-Party Software Components Present Cybersecurity Challenges
For a number of reasons, the incorporation of third-party software into medical devices creates challenges for cybersecurity. If a medical device is compromised due to software vulnerability, it could disrupt patient care, perhaps at a systemwide level, or could lead to a data breach.
Among the cyber-attacks cited by ECRI are the WannaCry ransomware attack in 2017.
Artificial Intelligence (AI) Applications for Diagnostic Imaging May Misrepresent Certain Patient Populations
As AI begins to make an impact in healthcare, especially in diagnostic imaging, healthcare organizations should be aware of the limitations of current AI-based technologies.
The quality of the conclusions reached by AI algorithms, ECRI points out, depends on the quality of the data used to train the AI application. “Unreliable AI functionality can lead to misdiagnoses or can prompt inappropriate care decisions.”
A key challenge in developing an AI algorithm is overcoming bias in the data, the report says. AI software is inherently biased toward patient populations that “look like” the population used in building the algorithm. If that data doesn’t accurately represent a particular population, the resultant output may not be appropriate for those patients.
Remote Operation of Medical Devices Designed For Bedside Use Introduces Insidious Risks
During the surge periods of the pandemic, methods for remotely operating ventilators, infusion pumps, and other devices have been deployed as a way to conserve PPE, minimize health worker exposure, and avoid delays associated with donning PPE. But when medical devices designed to be operated at the bedside are operated remotely instead, the switch can result in challenges for patient care. Among ECRI’s concerns:
Less frequent visual assessment of the patient, which may prevent staff from observing clinically relevant conditions or device complications
Adverse effects on device performance because of longer tubing sets or staff being unable to see or hear the functioning of the device
Infection risks associated with increased connection points on infusion tubing or with compromised patient isolation (eg, if cables are channeled through an ajar door)
Insufficient Quality Assurance of 3D-Printed Patient-Specific Medical Devices May Harm Patients
According to ECRI, “3D-printing technology is now being used to create a range of patient-specific devices, including implants, anatomical models for surgical planning, surgical guides for orthopedic procedures, and prostheses.”
However, the report notes, “the use of an improperly created 3D-printed device could lead to procedure delays, surgical complications, infection, or patient injury.”
ECRI throws in a couple of other caveats: “Considerable engineering expertise” is required to convert imaging data into a digital design and make the object on a 3D printer. Also, the physician who will be using a patient-specific 3D-printed device plays a key role in the design process. So doctors bear increased responsibility for making sure that quality assurance measures have been followed in creating a 3D-printed object.