Clinical laboratories involved in genetic testing may find this welcomed news, after a pair of studies conducted in 2019 raised concerns about CRISPR base editing. The researchers of those studies observed that it “causes a high number of unpredictable mutations in mouse embryos and rice,” Chemical and Engineering News (C&EN) reported, adding, “Other groups have raised concerns about off-target mutations caused when the traditional CRISPR-Cas9 form of gene editing cuts DNA at a location that it wasn’t supposed to touch. The results of the new studies are surprising, however, because scientists have lauded base editors as one of the most precise forms of gene editing yet.”
Nevertheless, UC Berkeley’s latest breakthrough is expected to drive development of new and more accurate CRISPR-Cas genome-editing tools, which consist of base editors as well as nucleases, transposases, recombinases, and prime editors.
Understanding CRISPR Base Editors At a ‘Deeper Level’
Harvard University Chemistry and Chemical Biology Professor David Liu, PhD, who co-authored the study, explained the significance of this latest discovery.
“While base editors are now widely used to introduce precise changes in organisms ranging from bacteria to plants to primates, no one has previously observed the three-dimensional molecular structure of a base editor,” he said in a UC Berkeley news release. “This collaborative project reveals the beautiful molecular structure of a state-of-the-art highly-active base editor—ABE8e—caught in the act of engaging a target DNA site.”
UC Berkeley Professor Jennifer Doudna, PhD (above), who served as senior author of the study, says scientists may now have the information necessary to refine base editors and improve their precision and genome-targeting ability. “This structure helps us understand base editors at a much deeper level,” she said in the UC Berkeley statement. “Now that we can see what we’re working with, we can develop informed strategies to improve the system.” (Photo copyright: UC Berkeley.)
Jennifer Doudna, PhD, UC Berkeley Professor, Howard Hughes Medical Institute Investigator, and senior author of the study, has been a leading figure in the development of CRISPR-Cas9 gene editing. In 2012, Doudna and Emmanuelle Charpentier, PhD, Founding, Scientific and Managing Director at Max Planck Unit for the Science of Pathogens in Berlin, led a team of researchers who “determined how a bacterial immune system known as CRISPR-Cas9 is able to cut DNA, and then engineered CRISPR-Cas9 to be used as a powerful gene editing technology.” In a 2017 news release, UC Berkeley noted that the work has been described as the “scientific breakthrough of the century.”
Viewing the Base Editor’s 3D Shape
CRISPR-Cas9 gene editing allows scientists to permanently edit the genetic information of any organism, including human cells, and has been used in agriculture as well as medicine. A base editor is a tool that manipulates a gene by binding to DNA and replacing one nucleotide with another.
According to the recent UC Berkeley news release, the research team used a “high-powered imaging technique called cryo-electron microscopy” to reveal the base editor’s 3D shape.
Genetic Engineering and Biotechnology News notes that, “The high-resolution structure is of ABE8e bound to DNA, in which the target adenine is replaced with an analog designed to trap the catalytic conformation. The structure, together with kinetic data comparing ABE8e to earlier ABEs [adenine base editors], explains how ABE8e edits DNA bases and could inform future base-editor design.”
The graphic above, taken from the UC Berkeley news release, shows the “3D structure of a base editor, comprised of the Cas9 protein (white and gray), which binds to a DNA target (teal and blue helix) complementary to the RNA guide (purple), and the deaminase proteins (red and pink), which switch out one nucleotide for another.” (Image and caption copyright: UC Berkeley.)
Knowing the Cas9 fusion protein’s 3D structure may help eliminate unintended off-target effects on RNA, extending beyond the targeted DNA. However, until now, scientists have been hampered by their inability to understand the base editor’s structure.
“If you really want to design truly specific fusion protein, you have to find a way to make the catalytic domain more a part of Cas9, so that it would sense when Cas9 is on the correct target and only then get activated, instead of being active all the time,” study co-first author Audrone Lapinaite, PhD, said in the news release. At the time of the study, Lapinaite was a postdoctoral fellow at UC Berkeley. She is now an assistant professor at Arizona State University.
“As a structural biologist, I really want to look at a molecule and think about ways to rationally improve it. This structure and accompanying biochemistry really give us that power,” added UC Berkeley postdoctoral fellow Gavin Knott, PhD, another study co-author. “We can now make rational predications for how this system will behave in a cell, because we can see it and predict how it’s going to break or predict ways to make it better.”
Clinical laboratory leaders and pathologists will want to monitor these new advances in CRISPR technology. Each breakthrough has the power to fuel development of cost-effective, rapid point-of-care diagnostics.
The KCL researchers’ new models for predicting which patients will need hospitalization and breathing support may be useful for pathologists and clinical laboratory scientists
One more window into understanding the SARS-CoV-2 coronavirus may have just opened. A British study identified six distinct “clusters” of symptoms that the research scientists believe may help predict which patients diagnosed with COVID-19 will require hospitalization and respiratory support. If further research confirms these early findings, pathologists and medical laboratory managers may gain new tools to diagnose infections faster and more accurately.
Launched in March in the United Kingdom and extended to the United States and Sweden, the app has attracted more than four million users who track their health and potential COVID symptoms on a daily basis.
Increased Accuracy in Predicting COVID-19 Hospitalizations
KCL researchers identified six distinct “types” of COVID-19, each distinguished by a particular cluster of symptoms. They include headaches, muscle pains, fatigue, diarrhea, confusion, loss of appetite, shortness of breath, and more. The researchers also found that COVID-19 disease progression and outcome also vary significantly between people, ranging from mild flu-like symptoms or a simple rash to severe or fatal conditions.
Using app data logged by 1,600 users in March and April, the researchers developed an algorithm that combined information on age, gender, body mass index (BMI), and pre-existing conditions with recorded symptoms from the onset of the illness through the first five days. The researchers then tested the algorithm using a second independent dataset of 1,000 users, logged in May.
In a news release, the KCL researchers identified the six clusters of symptoms as:
Flu-like with No Fever: Headache, loss of smell, muscle pains, cough, sore throat, chest pain, no fever.
Flu-like with Fever: Headache, loss of smell, cough, sore throat, hoarseness, fever, loss of appetite.
Gastrointestinal: Headache, loss of smell, loss of appetite, diarrhea, sore throat, chest pain, no cough.
Severe Level One, Fatigue: Headache, loss of smell, cough, fever, hoarseness, chest pain, fatigue.
Severe Level Two, Confusion: Headache, loss of smell, loss of appetite, cough, fever, hoarseness, sore throat, chest pain, fatigue, confusion, muscle pain.
Severe Level Three, Abdominal and Respiratory: Headache, loss of smell, loss of appetite, cough, fever, hoarseness, sore throat, chest pain, fatigue, confusion, muscle pain, shortness of breath, diarrhea, abdominal pain.
Using the data, the researchers were able to more accurately predict—78.8% versus 69.5%—which of the six symptom clusters placed patients at higher risk of requiring hospitalization and breathing support (ventilation or additional oxygen) than with prediction models based on personal characteristics alone. For example, nearly 50% of the patients in cluster six (Severe Level Three, Abdominal and Respiratory) ended up in the hospital, compared with 16% of those in cluster one (Flu-like with No Fever).
“These findings have important implications for care and monitoring of people who are most vulnerable to severe COVID-19,” Claire Steves, MD, PhD (above left), Clinical Senior Lecturer at King’s College London, said in the KCL news release. “If you can predict who these people are at day five, you have time to give them support and early interventions, such as monitoring blood oxygen and sugar levels, and ensuring they are properly hydrated—simple care that could be given at home, preventing hospitalizations and saving lives.” (Photo copyright: King’s College London.)
According to the Zoe website, the ongoing research is led by:
Prof. Tim Spector, FMedSci, Professor of Genetic Epidemiology at King’s College London and Director of TwinsUK, an adult registry of twins in the United Kingdom;
Andrew Chan, MD, Professor of Medicine at Harvard Medical School, Professor of Immunology and Infectious Diseases, Harvard T.H. Chan School of Public Health and Chief of the Clinical and Translational Epidemiology Unit, CTEU Massachusetts General Hospital; and
Encouraging Everyone to Use the COVID-Symptom Study App
The study points out that—broadly speaking—people with cluster four, five, or six COVID-19 symptoms tended to be older and frailer and were more likely to be overweight and have pre-existing conditions, such as diabetes or lung disease, than those with cluster one, two, or three symptoms.
“Our study illustrates the importance of monitoring symptoms over time to make our predictions about individual risk and outcomes more sophisticated and accurate,” said lead researcher Carole Sudre, PhD (above), a Research Fellow at King’s College London and the study’s lead researcher, in the KCL news release. “This approach is helping us to understand the unfolding story of this disease in each patient so they can get the best care.” (Photo copyright: University College London.)
Tim Spector, FMedSci, Head of the Department of Twin Research and Genetic Epidemiology, and Professor of Genetic Epidemiology at King’s College London, encourages everyone to download the COVID Symptom Study app and help increase the data available to researchers.
“Data is our most powerful tool in the fight against COVID-19,” Spector said in the KCL news release. “We urge everyone to get in the habit of using the app daily to log their health over the coming months, helping us to stay ahead of any local hotspots or a second wave of infections.”
As the body of knowledge surrounding COVID-19 grows, clinical laboratory professionals would be well advised to remain informed on further research regarding not only the potential for COVID-19 variants to exist, but also the evolving guidance on infection prevention and testing.
In the absence of a “gold standard,” researchers are finding a high frequency of false negatives among SARS-CoV-2 RT-PCR tests
Serology tests designed to detect antibodies to the SARS-CoV-2 coronavirus that causes the COVID-19 illness have been dogged by well-publicized questions about accuracy. However, researchers also are raising concerns about the accuracy of molecular diagnostics which claim to detect the actual presence of the coronavirus itself.
“Diagnostic tests, typically involving a nasopharyngeal swab, can be inaccurate in two ways,” said Steven Woloshin, MD, MS, in a news release announcing a new report that “examines challenges and implications of false-negative COVID-19 tests.” Woloshin is an internist, a professor at Dartmouth Institute, and co-director of the Geisel School of Medicine at Dartmouth.
“A false-positive result mistakenly labels a person infected, with consequences including unnecessary quarantine and contact tracing,” he stated in the news release. “False-negative results are far more consequential, because infected persons who might be asymptomatic may not be isolated and can infect others.”
Woloshin led a team of Dartmouth researchers who analyzed two studies from Wuhan, China, and a literature review by researchers in Europe and South America that indicated diagnostic tests for COVID-19 are frequently generating false negatives. The team published their results in the June 5 New England Journal of Medicine (NEJM).
For example, one research team in Wuhan collected samples from 213 hospitalized COVID-19 patients and found that an approved RT-PCR test produced false negatives in 11% of sputum samples, 27% of nasal samples, and 40% of throat samples. Their research was published on the medRxiv preprint server and has not been peer-reviewed.
The literature review Woloshin’s team studied was also published on medRxiv, titled, “False-Negative Results of Initial Rt-PCR Assays for COVID-19: A Systematic Review.” It indicated that the rate of false negatives could be as high as 29%. The authors of the review looked at five studies that had enrolled a total of 957 patients. “The collected evidence has several limitations, including risk of bias issues, high heterogeneity, and concerns about its applicability,” they wrote. “Nonetheless, our findings reinforce the need for repeated testing in patients with suspicion of SARS-Cov-2 infection.”
Another literature review, published in the Annals of Internal Medicine, titled, “Variation in False-Negative Rate of Reverse Transcriptase Polymerase Chain Reaction–Based SARS-CoV-2 Tests by Time Since Exposure,” estimated the probability of false negatives in RT-PCR tests at varying intervals from the time of exposure and symptom onset. For example, the authors found that the median false-negative rate was 38% if a test was performed on the day of symptom onset, versus 20% three days after onset. Their analysis was based on seven studies, five of which were peer-reviewed, with a total of 1330 test samples.
Doctors also are seeing anecdotal evidence of false negatives. For example, clinicians at UC San Diego Health medical center treated a patient with obvious symptoms of COVID-19, but two tests performed on throat samples were negative. However, a third test, using a sample from a bronchial wash, identified the virus, reported Medscape.
Sensitivity and Specificity of COVID-19 Clinical Laboratory Tests
The key measures of test accuracy are sensitivity, which refers to the ability to detect the presence of the virus, and specificity, the ability to determine that the targeted pathogen is not present. “So, a sensitive test is less likely to provide a false-negative result and a specific test is less likely to provide a false-positive result,” wrote Kirsten Meek, PhD, medical writer and editor, in an article for ARUP Laboratories.
“Analytic” sensitivity and specificity “represent the accuracy of a test under ideal conditions in which specimens have been collected from patients with either high viral loads or a complete absence of exposure,” she wrote. However, “sensitivity and specificity under real-world conditions, in which patients are more variable and specimen collection may not be ideal, can often be lower than reported numbers.”
In a statement defending its ID Now molecular point-of-care test, which came under scrutiny during a study of COVID-19 molecular tests by NYU Langone Health, Northwell Health, and Cleveland Clinic, according to MedTech Dive, Abbott Laboratories blamed improper sample collection and handling for highly-publicized false negatives produced by its rapid test. An FDA issued alert about the test on May 14 noted that Abbott had agreed to conduct post-market studies to identify the cause of the false negatives and suggest remedial actions.
Issues with Emergency Use Authorizations
In their NEJM analysis, Woloshin et al point to issues with the FDA’s process for issuing Emergency Use Authorizations (EUAs). For example, they noted variations in how manufacturers are conducting clinical evaluations to determine test performance. “The FDA prefers the use of ‘natural clinical specimens’ but has permitted the use of ‘contrived specimens’ produced by adding viral RNA or inactivated virus to leftover clinical material,” they wrote.
When evaluating clinical performance, manufacturers ordinarily conduct an index test of patients and compare the results with reference-standard test, according to the Dartmouth researchers. For people showing symptoms, the reference standard should be a clinical diagnosis performed by an independent adjudication panel. However, they wrote, “it is unclear whether the sensitivity of any FDA-authorized commercial test has been assessed in this way.” Additionally, a reference standard for determining sensitivity in asymptomatic people “is an unsolved problem that needs urgent attention to increase confidence in test results for contact-tracing or screening purposes.”
“To truly determine false negatives, you need a gold standard test, which is essentially as close to perfect as we can get,” Stephen Rawlings, MD, PhD, (above), a resident physician of internal medicine and infectious diseases fellow at UC San Diego’s Center for AIDS Research (CFAR), who has been working on SARS-CoV-2 test validation since March. “But there just isn’t one yet for coronavirus,” he told Medscape. (Photo copyright: University of California, San Diego.)
Continued adherence to current measures, such as physical distancing and surface disinfection.
Development of highly sensitive and specific tests or combinations of tests to minimize the risk of false-negative results and ongoing transmission based on a false sense of security.
Improved RT-PCR tests and serological assays.
Development and communication of clear risk-stratified protocols for management of negative COVID-19 test results.
“These protocols must evolve as diagnostic test, transmission, and outcome statistics become more available,” they wrote.
Meanwhile, clinical laboratories remain somewhat on their own at selecting which COVID-19 molecular and serology tests they want to purchase and run in their labs. Complicating such decisions is the fact that many of the nation’s most reputable in vitro diagnostics manufacturers cannot produce enough of their COVID-19 tests to meet demand.
Consequently, when looking to purchase tests for SARS-CoV-2, smaller medical laboratory organizations find themselves evaluating COVID-19 kits developed by little-known or even brand-new companies.
Even as some states lift stay-at-home orders, clinical laboratories and pathology groups face uncertainty about how quickly routine daily test referrals will return to normal, pre-pandemic levels
Although strokes and heart attacks do not take vacations, a large and growing number of patients with serious health issues who—in normal times—would require immediate attention are not contacting providers to get needed care. Instead, they are avoiding hospital emergency rooms and clinical laboratories for fear they’ll contract the COVID-19 coronavirus.
Starting in early March, hospitals nationwide suspended elective surgeries and procedures and reduced non-COVID-19 inpatient care to make beds available for the predicted on-rush of COVID-19 patients. However, in parts of the country, the predicted high demand for hospital beds and ventilators failed to materialize. Additionally, due to shelter-in-place orders, patients in many states postponed routine office visits with their primary care physicians.
The collective collapse in the number of elective services provided by hospitals, and the fall-off in patients visiting their doctors, is crushing the financial stability of the nation’s clinical laboratory industry.
In, “From Mid-March, Labs Saw Big Drop in Revenue,” Dark Daily’s sister publication, The Dark Report (TDR) reported on the revenue challenges facing clinical pathology groups and clinical laboratories. Kyle Fetter, Executive Vice President and General Manager of Diagnostic Services at XIFIN, a revenue cycle management company, told TDR that starting in the third week of March, labs suffered a steep decline in routine testing. By the end of March, that fall-off in revenue ranged from 44% for some AP specimens to 70% to 80% for some specialty AP work. During these same weeks, XIFIN’s data showed clinical labs experienced a drop in routine testing volume of 58%, hospital outreach testing declined by 61%, and molecular lab volume went down by 52%.
Can Clinical Laboratories Hang on Financially Until COVID-19 Goes Away?
Though most states have not met the nonbinding criteria recommended by the Trump administration for reopening, nearly 40 governors in early May began loosening stay-at-home orders, reported CNN, including allowing elective medical procedures to resume.
Patients may make up for lost time by returning to doctors’ offices for medical laboratory tests and other COVID-19-delayed procedures, and as this happens, clinical laboratories may experience a surge in routine test orders from doctors’ offices and hospital admissions once stay-at-home orders are lifted and fear of COVID-19 has passed.
According to an article published on Axios, a survey of 163 physicians conducted by SVB Leerink—an investment firm that specializes in healthcare and life sciences—found that “roughly three out of four doctors believe patient appointments will resume to normal, pre-coronavirus levels, no earlier than July, and 45% expect a rebound to occur sometime between July and September.” If so, the financial squeeze facing clinical laboratories, pathology groups, and other medical and dental professionals may continue to loosen.
Christopher Freer, DO (above), an emergency physician at St. Barnabas Hospital in the Bronx and Director of Emergency Medicine at RWJBarnabas Health, told CNBC that emergency departments are seeing patients with severe issues, such as stroke and appendicitis, but that those with milder symptoms appear to be staying away. “Even with coronavirus, we still have healthy people who get an illness and need to go to the emergency room,” he said. “Heart attacks don’t stop.” (Photo copyright: USA Today.)
Hospital Finances Are Being Particularly Stressed by Loss of Patients
The impact of stay-at-home orders on hospital systems, in particular, has been dramatic. CNBC reported that RWJBarnabas Health, an 1l-hospital 22-laboratory health system in New Jersey that has 11 emergency departments, totaled just 180 emergency room visits per day during a mid-April weekend, a sharp decline from their 280-per-day-average.
A recent Washington Post article paints an even bleaker picture. Clinicians in the United States, Spain, United Kingdom, and China anecdotally report a “silent sub-epidemic of people who need care at hospitals but dare not come in,” the article states, noting people with symptoms of appendicitis, heart attacks, stroke, infected gall bladders, and bowel obstructions are avoiding hospital emergency rooms.
“Everybody is frightened to come to the ER,” Mount Sinai Health System cardiovascular surgeon John Puskas, MD, told the Post. Though his 60-bed cardiac unit had been repurposed to care for COVID-19 patients, Puskas said the New York hospital system was seeing “dramatically fewer” cardiac patients.
Concerned that patients may be ignoring signs of heart attack or stroke rather than go to a hospital, the American College of Cardiology launched the “CardioSmart” campaign, which urges anyone experiencing heart symptoms to get prompt treatment and to continue routine appointments, using telehealth technology when available.
“Hospitals have safety measures to protect you from infection,” the CardioSmart website states. “Getting care quickly is critical. You’ll get better faster, and you’ll limit damage to your health.”
However, David Brown, MD, Chief of Emergency Medicine at Massachusetts General Hospital in Boston, argues the number of people having heart-related issues is unlikely to have dropped during the pandemic.
“Strokes and heart attacks don’t take a vacation just because there’s a pandemic,” Brown told The Boston Globe. “They’re still happening. They just aren’t happening as much inside the hospital, which is a major concern to me.”
Many healthcare professionals are worried about the long-term effect from pandemic-delayed preventative and elective procedures.
“The big question is are we going to see a lot more people that have bad outcomes from heart disease, from stroke, from cancer because they’ve put off what they should have had done, but were too afraid to come to the hospital?” Providence St. Joseph Health CEO Rod Hochman, MD, told CNBC.
Hochman, who is Chair-elect of the American Hospital Association (AHA), maintains the aftereffects of people putting off elective surgeries and screening procedures like colonoscopies and mammograms may be felt for years to come.
“We’re possibly going to see a blip in other disease entities as a consequence of doubling down on COVID-19,” he told CNBC.
In clinical laboratories, COVID-19 testing may have somewhat helped offset the drop in routine testing volume. However, the pandemic’s overall financial costs to labs and pathology groups will likely be felt for months to years, as patients slowly return to healthcare providers’ offices and hospitals.
Researchers find a savings of more than one million dollars and prevention of hundreds, if not thousands, of adverse drug events could have been had with machine learning system
Support for artificial intelligence (AI) and machine learning (ML) in healthcare has been mixed among anatomic pathologists and clinical laboratory leaders. Nevertheless, there’s increasing evidence that diagnostic systems based on AI and ML can be as accurate or more accurate at detecting disease than systems without them.
Dark Daily has covered the development of artificial intelligence and machine learning systems and their ability to accurately detect disease in many e-briefings over the years. Now, a recent study conducted at Brigham and Women’s Hospital (BWH) and Massachusetts General Hospital (MGH) suggests machine learning can be more accurate than existing clinical decision support (CDS) systems at detecting prescription medication errors as well.
The study was partially retrospective in that the
researchers compiled past alerts generated by the CDS systems at BWH and MGH
between 2009-2011 and added them to alerts generated during the active part of
the study, which took place from January 1, 2012 to December 31, 2013, for a
total of five years’ worth of CDS alerts.
They then sent the same patient-encounter data that generated those CDS alerts to a machine learning platform called MedAware, an AI-enabled software system developed in Ra’anana, Israel.
MedAware was created for the “identification and prevention
of prescription errors and adverse drug effects,” notes the study, which goes
on to state, “This system identifies medication issues based on machine
learning using a set of algorithms with different complexity levels, ranging
from statistical analysis to deep learning with neural networks. Different
algorithms are used for different types of medication errors. The data elements
used by the algorithms include demographics, encounters, lab test results,
vital signs, medications, diagnosis, and procedures.”
The researchers then compared the alerts produced by
MedAware to the existing CDS alerts from that 5-year period. The results were
astonishing.
According to the study:
“68.2% of the alerts generated were unique to
the MedAware system and not generated by the institutions’ CDS alerting system.
“Clinical outlier alerts were the type least
likely to be generated by the institutions’ CDS—99.2% of these alerts were
unique to the MedAware system.
“The largest overlap was with dosage alerts,
with only 10.6% unique to the MedAware system.
“68% of the time-dependent alerts were unique to
the MedAware system.”
Perhaps even more important was the results of the cost
analysis, which found:
“The average cost of an adverse event
potentially prevented by an alert was $60.67 (range: $5.95–$115.40).
“The average adverse event cost per type of
alert varied from $14.58 (range: $2.99–$26.18) for dosage outliers to $19.14
(range: $1.86–$36.41) for clinical outliers and $66.47 (range: $6.47–$126.47)
for time-dependent alerts.”
The researchers concluded that, “Potential savings of $60.67 per alert was mainly derived from the prevention of ADEs [adverse drug events]. The prevention of ADEs could result in savings of $60.63 per alert, representing 99.93% of the total potential savings. Potential savings related to averted calls between pharmacists and clinicians could save an average of $0.047 per alert, representing 0.08% of the total potential savings.
“Extrapolating the results of the analysis to the 747,985
BWH and MGH patients who had at least one outpatient encounter during the
two-year study period from 2012 to 2013, the alerts that would have been fired
over five years of their clinical care by the machine learning medication
errors identification system could have resulted in potential savings of
$1,294,457.”
Savings of more than one million dollars plus the prevention
of potential patient harm or deaths caused by thousands of adverse drug events
is a strong argument for machine learning platforms in diagnostics and
prescription drug monitoring.
Researchers Say Current Clinical Decision Support Systems
are Limited
Machine learning is not the same as artificial intelligence. ML is a “discipline of AI” which aims for “enhancing accuracy,” while AI’s objective is “increasing probability of success,” explained Tech Differences.
Healthcare needs the help. Prescription medication errors cause patient harm or deaths that cost more than $20 billion annually, states a Joint Commission news release.
CDS alerting systems are widely used to improve patient
safety and quality of care. However, the BWH-MGH researchers say the current
CDS systems “have a variety of limitations.” According to the study:
“One limitation is that current CDS systems are rule-based and can thus identify only the medication errors that have been previously identified and programmed into their alerting logic.
“Further, most have high alerting rates with many false positives, resulting in alert fatigue.”
Commenting on the value of adding machine learning
medication alerts software to existing CDS hospital systems, the BWH-MGH
researchers wrote, “This kind of approach can complement traditional rule-based
decision support, because it is likely to find additional errors that would not
be identified by usual rule-based approaches.”
However, they concluded, “The true value of such alerts is
highly contingent on whether and how clinicians respond to such alerts and
their potential to prevent actual patient harm.”
Future research based on real-time data is needed before machine
learning systems will be ready for use in clinical settings, HealthITAnalytics
noted.
However, medical laboratory leaders and pathologists will
want to keep an eye on developments in machine learning and artificial
intelligence that help physicians reduce medication errors and adverse drug
events. Implementation of AI-ML systems in healthcare will certainly affect
clinical laboratory workflows.
