A New York Times report suggests that frequent testing is still the best approach to controlling spread of the SARS-CoV-2 coronavirus
Many colleges and universities go to great lengths to screen their students for signs of COVID-19 using technologies that include fever scanners, heart-rate monitors, and symptom-checking apps. But a recent report in The New York Times, titled, “Colleges That Require Virus-Screening Tech Struggle to Say Whether It Works,” suggests that academic institutions would be better off adopting frequent clinical laboratory testing for the SARS-CoV-2 coronavirus, even if it is more expensive than symptom screening.
This shouldn’t be a surprise to pathologists and other medical laboratory professionals who have followed news and research about the pandemic. Back in Sept. 2020, the federal Centers for Disease Control and Prevention (CDC) in a media statement noted that “symptom-based screening has limited effectiveness because people with COVID-19 may have no symptoms or fever at the time of screening, or only mild symptoms.”
That same month, Medscape reported that presidential advisor Anthony Fauci, MD, said, “It is now clear that about 40%-45% of infections are asymptomatic.”
But this hasn’t prevented educational institutions from investing in costly screening technologies. One cited by The New York Times (NYT) was the University of Idaho, where 9,000 students live on or near campus. The university has spent $90,000 on fever scanners resembling airport metal detectors, the paper reported, but as of early March, the units had identified fewer than 10 people with high skin temperatures.
“Even then, university administrators could not say whether the technology had been effective because they have not tracked students flagged with fevers to see if they went on to get tested for the virus,” the NYT reported, adding that many other institutions that adopted screening technologies have failed to systematically measure the effectiveness of these approaches.
“The moral of the story is you can’t just invest in this tech without having a validation process behind it,” infectious-disease epidemiologist Saskia Popescu PhD, MPH, of George Mason University told The New York Times.
Rising COVID-19 Infections on College Campuses
These efforts have come amid increasing COVID-19 infection rates on many US campuses. In “Cases Rise, Restrictions Begin,” Inside Higher Ed reported that large universities were doing better than they had in the fall 2020 semester, but that “other campuses—including those that kept cases low in the fall—are seeing numbers rise.” One such campus was Boston College, which cast blame on students who were not following safety protocols.
For its story, The New York Times surveyed more than 1,900 US colleges and universities as part of an effort to track outbreaks on campus. Respondents reported more than 120,000 campus-related COVID-19 cases between Jan. 1 and March 2, 2021, but because institutions measure outbreaks in different ways, the NYT reported that this is likely an undercount. Overall, institutions reported more than 535,000 cases since the pandemic began, according to the survey.
Clinical Laboratory Testing Still Ongoing on College Campuses
School administrators told The New York Times that despite questions about the usefulness of screening tools, this approach is still worthwhile as reminders for students to follow other protocols, such as mask wearing.
And universities have not abandoned testing for COVID-19. For example, The New York Times noted that students at the University of Idaho are tested at least twice each semester, and the school is also testing wastewater to identify outbreaks of SARS-CoV-2.
The Ohio State News, a publication of Ohio State University, reported in late February that it had tested 30,000 people in a single week, accounting for 12% of the COVID-19 tests conducted in Ohio. At the start of the fall semester, the university was sending test samples to a private company in New Jersey, but later it began processing samples at the on-campus Applied Microbiology Services Lab (AMSL).
“By the start of spring semester, the AMSL was processing about 85% of Ohio State’s COVID-19 tests,” the university reported, for a likely savings of $30 million to $40 million. Leaders of the testing program expect that they can realistically conduct 35,000 tests per week.
Chris Marsicano, PhD (above), a professor and researcher at Davidson College, told Inside Higher Ed that many institutions are relying on antigen testing, which is less costly but also less reliable than PCR (polymerase chain reaction) tests. “PCR tests are expensive,” he said. “Just because you’re testing multiple times a week doesn’t mean you’re catching all the cases.” Marsicano leads the institution’s College Crisis Initiative. Clinical laboratory leaders can attest to Marsicano’s statement. (Photo copyright: Twitter.)
Using Technology for COVID-19 Contact Tracing
In addition to symptom screening, some universities have adopted technologies that track student movement on campus for contact-tracing purposes. But again, the benefits are questionable. For example, Bridgewater State University in Bridgewater, Mass. asked students to scan QR codes at various locations, but only one-third were doing so, The New York Times reported. Another system at the university records entry to campus buildings when students swipe their IDs.
“We found what we need is tests and more tests,” clinical psychologist Christopher Frazer, Psy.D., Executive Director of the university’s wellness center, told The New York Times. He said that students on campus are tested once a week. When they have tested positive, contact tracers “often learned much more about infected students’ activities by calling them than by examining their location logs,” the NYT reported.
Colleges and universities are also banking on vaccination to reduce the spread of the virus, Inside Higher Ed reported. Some will require all students to be vaccinated for the fall semester, but such mandates are facing legal and political hurdles. For example, executive orders by Texas Governor Greg Abbott and Florida Governor Ron DeSantis may prohibit institutions in those states from imposing vaccination requirements.
As colleges and universities struggle to deal with the challenges of COVID-19, clinical laboratories have resources for staying up to date on current testing and tracking technologies in use on campuses. For example, the CDC is funding a program to facilitate sharing of best practices and other information. Inside Higher Ed reported that the Higher Education COVID-19 Community of Practice (CoP) will include a discussion board, webinars, and a searchable database of info uploaded by participating institutions.
But information blocking remains a barrier to complete information exchange, creating ongoing issues for clinical laboratories and pathology groups
Interoperability of electronic health records (EHRs) remains one the biggest challenges for clinical laboratories and anatomic pathology groups that must interface with the EHRs of their physician clients to enable electronic transmission of medical laboratory orders and test results.
Laboratory professionals will be pleased to know the most recent federal government report on hospital interoperability shows 55% of all hospitals can now send, receive, find, and integrate patient information from outside sources into their EHRs. This is an important milestone on the road to robust data exchange.
About 70% of hospitals reported integrating data into their EHR—a nearly 15% increase from 2018.
A majority of hospitals used a mix of electronic and non-electronic methods to exchange summary of care records. However, use of electronic third-party methods, Health Information Service Providers (HISPs), health information exchange (HIE), and vendor networks increased in 2019.
The proportion of hospitals that used a national network to find (or query) patient health information increased by nearly 40% between 2018 and 2019.
The graphic above taken from the ONC report shows the “percent of US non-federal acute care hospitals that electronically find patient health information and send, receive, and integrate patient summary of care records from sources outside their health system from 2014-2019. About 70% of hospitals reported integrating data into their EHR—a nearly 15% increase from 2018.” This is a positive development for clinical laboratories and anatomic pathology groups, because it makes it easier for them to accept electronic medical laboratory test orders and report test results electronically. (Graphic copyright: Office of the National Coordinator of Health Information Technology.)
David Burda, creator and leader of 4sight Health, a thought leadership and advisory company, has been a forceful advocate for healthcare interoperability, routinely stressing that patients cannot receive the optimum level of care from their providers as long as EHR vendors and health systems engage in information blocking.
In a blog post, Burda commented on the ONC report and outlined how far there still is to go. “Hospitals passed an important interoperability milestone in 2019, but the goal of reaching total hospital interoperability is still ways off.
“To be fair,” he added, “there were some other signs of progress in the new ONC report. The most significant, from a patient’s point of view, was the fact that in 2019, more hospitals were actively seeking patient health information from other providers and sources as part of how they routinely diagnose and treat patients. They’re not passively relying on data in their own EHR systems to make medical decisions.”
For example, Burda wrote:
73% of the hospitals said they struggle with exchanging patient information with other providers who use a different EHR system.
66% of the hospitals said they share patient information with other providers who don’t share patient health information with them.
59% of the hospitals said other providers’ EHR systems don’t have the capability to receive patient health information from them.
“These [issues] are all caused by cultural, financial, and technical barriers that should have fallen years ago,” wrote David Burda (above), news editor and columnist for 4sight Health in his blog post about the ONC interoperability report. “But they didn’t, and all we can do is keep pushing forward to the day patient health data stops being a closely guarded commodity and starts flowing freely throughout the delivery system to drive better care for patients.” Clinical laboratory test results, being the largest portion of data contained in electronic health records, would make up a significant portion of the health data Burda is referring to. (Photo copyright: 4sight Health.)
KLAS and CHIME are Optimistic about EHR Interoperability
Industry progress toward interoperability was also noted in a white paper titled, “Trends in EMR Interoperability,” co-authored by KLAS Research and the College of Healthcare Information Management Executives (CHIME). The authors found reasons for optimism, noting the rate of provider organizations achieving “deep interoperability” had doubled since 2017, with roughly two-thirds of provider organizations often or nearly always having access to needed records.
“The overall rate leaves much to be desired, but signs of progress are visible,” the authors wrote. Evidence of that progress includes improved data sharing with outside EHRs, a growing ability for ambulatory clinics and smaller hospitals to connect with larger organizations, and more widespread use of national networks to achieve information sharing.
“Since KLAS’ prior large-scale interoperability study in 2017, the market has made notable progress; access to outside records has increased, provider organizations are connecting to more critical exchange partners than ever, and the use of APIs offers new ways to facilitate data exchange in service of myriad use cases,” the report concludes. “Even with all this progress, there is still a significant opportunity for EMR (electronic medical record) vendors and provider organizations to partner effectively to help data exchange truly impact patient care. With additional work, the industry appears poised for improvement in this area going forward.”
Seema Verma says Interoperability is Improving
In an article she authored for Health IT News, titled, “How CMS Has Made Progress on Healthcare Interoperability,” Seema Verma, Administrator for the Centers for Medicare and Medicaid Services (CMS) during the Trump presidency, noted that great strides have been made in recent years toward the goal of complete interoperability.
“Technology is ever evolving, and our work will constantly evolve, but our efforts have laid a foundation for future policy that will enable the secure and interoperable exchange of healthcare information, drive value-based care in America, and give patients and doctors the information they need,” she wrote.
For clinical laboratories and anatomic pathology groups, the road to interoperability remains littered with a few potholes, but speed bumps are disappearing, which may signal a time in the not-too-distant future when clinical laboratories and pathology groups will easily interface electronically with physicians, hospitals, and other providers to receive test orders and transmit test results.
COVID-19 has made telehealth an important tool. New technologies may help clinical laboratories collect blood samples ordered by physicians treating patients remotely
Even before COVID-19, telehealth services were gaining in popularity. But the SARS-CoV-2 pandemic fully opened the door to widespread use of mobile healthcare (mHealth) technologies. This has had an on-going impact on clinical laboratories.
Pre-pandemic, if a patient visited a healthcare provider and that provider ordered medical laboratory tests, the patient could simply walk down the hall to the lab’s patient service center and provide a blood sample. But when patients and providers meet through telehealth services, it is not so easy for lab personnel to collect samples for testing.
Several questions face healthcare providers and clinical laboratories as the pandemic subsides:
Will telehealth remain popular?
Does it benefit patient care?
Can physicians fit it into their workflows?
Will it continue to be reimbursed fairly?
COVID-19 Gives Telehealth Adoption a Big Boost
Telemedicine became important very quickly as SARS-CoV-2 coronavirus infections spread in early 2020. And not just in the United States. Clinicians worldwide began to embrace mHealth technology as a method of delivering care in a way that reduced the transmission of the virus.
The number of telemedicine consultations has declined since April 2020 but continues to be significantly higher than before the pandemic. It is also interesting to note that 90% of telemedicine visits were by phone in Australia and Canada, according to an article published in JAMA Network, titled, “Paying for Telemedicine After the Pandemic.”
“At its peak in April 2020, telemedicine was responsible for 38% of all ambulatory visits among Australia’s Medicare program, 42% of all ambulatory visits for individuals insured by a US commercial insurer, and 77% of all ambulatory visits among people in Ontario, Canada,” wrote Ateev Mehrotra, MD, MPH (above), Associate Professor of Health Care Policy, Harvard Medical School, and Associate Professor of Medicine and Hospitalist, Beth Israel Deaconess Medical Center (BIDMC), et al, in the Jama Network article. Clinical laboratory testing was part of all of that and continues to find its way in this new world of mobile healthcare. (Photo copyright: Managed Healthcare Executive.)
Telehealth Popular with Community Health Centers but Disparities Remain
One of the big issues with telehealth, according to the NACHC, is that not all patients have access to the technology necessary for telehealth to be a viable alternative to traditional office visits. And that patients who use NACHC clinics tend to be “low income, minority, and uninsured or publicly insured.”
Thus, the NACHC lists “inadequate broadband” as one of the biggest issues regarding the continued use of telehealth. “Patients without reliable internet or the necessary technology still face difficulties accessing services, which has resulted in forgone or delayed care,” the NACHC noted.
A study, titled, “Who Is (and Is Not) Receiving Telemedicine Care During the COVID-19 Pandemic,” published in the American Journal of Preventative Medicine (AJPM), confirms the findings of the NACHC. “The COVID-19 pandemic has affected telehealth utilization disproportionately based on patient age, and both county-level poverty rate and urbanicity.”
Although in-person visits declined by 50%, the AJPM study’s authors noted that telehealth did not completely bridge the gap, particularly in areas where there were higher levels of poverty.
Physician Practices Are Businesses Too
The pandemic hurt businesses of all types, including independent physician’s offices. Approximately 8% of practices closed due to the pandemic, and 4% expect they will shut down within the next year. Along with the financial burden of shutdowns, physicians are burning out, Fast Company reported.
Organizations now have the technology in place and some patients have learned to utilize the service. However, the situation does raise important questions:
Will telehealth remain a critical component of healthcare in the future?
As physician’s offices close, will telehealth fill the gap?
Telehealth and Payment
Becker’s Hospital Review asked nine hospital CIOs if telehealth would “have staying power.” Every executive mentioned either reimbursement or payers in their response. Therefore, whether telehealth remains a viable method of care delivery may depend more on who will pay for it and less on popularity or patient access.
During the COVID-19 pandemic, CMS revised the rules surrounding telehealth. This allowed practitioners to charge the same for telehealth visits as they would for in-person visits. Many private payers followed suit as well. However, those rules were temporary and it is not certain that they will be extended.
“Payers must continue to reimburse for telehealth visits,” Mark Amey, CIO, Alameda Health System, told Becker’s Hospital Review. “This has been approved with emergency orders, but there are questions on whether this will become permanent. The sooner this is addressed and resolved, the sooner organizations can make sure they are investing in permanent—not temporary—solutions.”
Tests that use nasal swabs and saliva have seen an enormous boom thanks to demand for COVID-19 testing that can be done at home, and COVID-19 antibody tests also are in high demand. Additionally, direct-to-consumer (DTC) tests that use blood samples also are seeing advancements. However, none of those factors—not even reimbursement—help medical laboratory managers who are trying to identify new methods of collecting specimens for testing that support telehealth doctors.
“Innovations in blood sample collection are proving their utility and validity just in time for the home-based medicine push,” noted the AACC. The article goes on to describe Mitra microsampling devices, produced by Neoteryx. These devices collect 20 uL of blood via a finger prick and are already used by organ transplant recipients.
Another method involves the use of dried blood spots.
Though COVID-19 is a factor, it is not the only one driving development of new healthcare technologies that may expand options for medical laboratories looking for ways to collect samples remotely.
As the COVID-19 pandemic progresses, we will continue to bring you news about healthcare technology that can enhance clinical laboratories’ ability to collect patient samples, include advancements in remote sampling techniques and technologies.
Oddly, as upcoding severity levels have risen, reported higher-severity inpatient hospital stays have dropped, OIG reported
Medicare upcoding fraud is a growing problem for the federal Centers for Medicare and Medicaid Services (CMS). Now, a report from the US Department of Health and Human Services (HHS) Office of Inspector General (OIG) suggests that the practice is increasingly occurring for high-severity inpatient hospital stays that account for the most expensive part of US healthcare.
“The [COVID-19] pandemic has placed unprecedented stress on the country’s healthcare system, making it more important than ever to ensure that Medicare dollars are spent appropriately,” the OIG report states.
The OIG website notes, “Medicare pays for many physician services using Evaluation and Management (commonly referred to as “E/M”) codes. New patient visits generally require more time than follow-up visits for established patients, and therefore E/M codes for new patients command higher reimbursement rates than E/M codes for established patients.”
The OIG describes one type of upcoding as “… an instance when [providers] provide a follow-up office visit or follow-up inpatient consultation, but bill using a higher-level E/M code as if [they] had provided a comprehensive new patient office visit or an initial inpatient consultation.
“Another example of upcoding related to E/M codes is misuse of Modifier 25,” the OIG continued. “Modifier 25 allows additional payment for a separate E/M service rendered on the same day as a procedure. Upcoding occurs if a provider uses Modifier 25 to claim payment for an E/M service when the patient care rendered was not significant, was not separately identifiable, and was not above and beyond the care usually associated with the procedure.”
How OIG Conducted the Study of Hospital Coding Practices
To perform its research, the OIG analyzed Medicare Part A claims for hospital stays for the six-year period from fiscal year (FY) 2014 through FY 2019. The OIG identified trends in billing and payments for inpatient hospital stays at the highest severity levels, as determined by the Medicare Severity Diagnosis Related Group (MS-DRG).
The OIG investigation revealed that the number of hospital stays billed at the highest severity level increased almost 20% between 2014 and 2019, while the number of stays billed at other severity levels decreased. These expenditures accounted for nearly half of all Medicare spending on inpatient hospital stays, the OIG reported.
According to the OIG report, “Medicare pays hospitals more for beneficiaries in MS-DRGs with higher severity levels because they are typically more costly to treat.” The graphic above taken from the OIG report illustrates “how the presence of complications can affect Medicare payment for three beneficiaries with the same principal diagnosis.” (Graphic copyright: Federal Office of Inspector General Department of Health and Human Services.)
As Severity Levels Went Up, Inpatient Length of Stays Went Down
Interestingly, the average length of inpatient stays at the highest severity level decreased, and the average length of hospital stays overall remained largely the same, decreasing by just 0.1 days. In addition, the total number of inpatient hospital stays decreased by 5%.
The OIG report noted that “the increase in the number of stays billed at the highest severity level implies that beneficiaries were sicker overall. However, the decrease in the average length of stays at the highest severity level potentially undermines that idea because it is not consistent with sicker beneficiaries. Length of stay generally has a positive relationship to severity of stay; sicker beneficiaries stay in the hospital longer.”
The OIG confirmed that in FY 2019, Medicare spent $109.8 billion for 8.7 million hospital stays. Approximately 3.5 million (or 40%) of those stays were billed at the highest severity level, as determined by the MS-DRG. In addition, nearly half of the $109.8 billion spent, or $54.6 billion, was for stays billed at the highest severity level and Medicare paid an average of $15,500 per stay at that level.
The OIG report states that “stays at the highest severity level are vulnerable to inappropriate billing practices, such as upcoding—the practice of billing at a level that is higher than warranted. Specifically, nearly a third of these stays lasted a particularly short amount of time and over half of the stays billed at the highest severity level had only one diagnosis qualifying them for payment at that level. Further, hospitals varied significantly in their billing of these stays, with some billing much differently than most.”
The OIG study also found that over half of the inpatient stays billed at the highest severity level achieved that level due to only one diagnosis. According to the OIG, the severity of an inpatient stay depends on a patient’s secondary diagnosis and it only takes one secondary diagnosis to propel a patient into the highest severity level. The OIG determined that if the diagnosis was inaccurate or inappropriate, higher payments would not be warranted.
OIG Recommends CMS Conduct Targeted Reviews
The report found that the most frequently billed MS-DRG in FY 2019 was septicemia or severe sepsis and that hospitals billed for 581,000 of these stays, for which Medicare paid $7.4 billion. In addition, kidney and urinary tract infections, pneumonia, and renal failure were among the most common conditions to have a complication that led to a high severity classification.
In its report, the OIG recommended more oversight from CMS to ensure that Medicare dollars are spent appropriately. The OIG also suggests that CMS conduct targeted reviews of MS-DRGs and hospital stays that are vulnerable to upcoding, as well as the hospitals that frequently bill them.
Clinical Laboratories Are Forewarned
Medicare audits continue to be more detailed and rigorous and all healthcare providers—including clinical laboratories and anatomic pathology groups—should be prepared to present all necessary documentation to support claims if and when they are audited.
Improvements in software, machine learning, and artificial intelligence (AI) give Medicare officials and the OIG powerful tools to spot questionable provider billing. This includes medical laboratories whose billing patterns could arouse suspicions and trigger audits.
Upcoding is a long-standing problem for the Medicare program. What is changing is that federal officials now have better tools and resources to use in identifying patterns of upcoding that fall outside accepted parameters.
By mining results of unrelated blood tests, the CIRRUS algorithm can inform doctors and patients earlier than usual of liver disease
For years Dark Daily and its sister publication The Dark Report have predicted that the same type of analytical software used on Wall Street to analyze bundles of debt, such as car loans, mortgages, and installment loans, would eventually find application in healthcare and clinical laboratory medicine. Now, researchers at the University of Southampton in England have developed just such an analytical tool.
The UK researchers call their algorithm CIRRUS, which stands for CIRRhosis Using Standard tests. It can, they say, accurately predict if a patient has cirrhosis of the liver at a much earlier stage than usual and produce information that is clinically actionable, using results from several common, routinely-ordered medical laboratory tests.
The University of Southampton scientists published their findings in BMJ Open.
Currently, the leading edge for this in clinical laboratory medicine is analysis of digital pathology images using image analysis tools and artificial intelligence (AI). However, CIRRUS is an example that analytical software is advancing in its ability to mine data from a number of clinically-unrelated lab tests on a patient and identify a health condition that might otherwise remain unknown.
The UK researchers designed the CIRRUS algorithm using routine clinical laboratory blood tests often requested in general practice to identify individuals at risk of advanced liver disease. These tests include:
“More than 80% of liver cirrhosis deaths are linked to alcohol or obesity and are potentially preventable,” noted Nick Sheron, MD, FRCP, Head of Population Hepatology at University of Southampton, and lead author of the study, in a press release. “However, the process of developing liver cirrhosis is silent and often completely unsuspected by GPs [general practitioners]. In 90% of these patients, the liver blood test that is performed is normal, and so liver disease is often excluded.
“This new CIRRUS algorithm can find a fingerprint for cirrhosis in the common blood tests done routinely by GPs,” he continued. “In most cases the data needed to find these patients already exists and we could give patients the information they need to change their lifestyle. Even at this late stage, if people address the cause by stopping drinking alcohol or reducing their weight, the liver can still recover.”
Mining Clinical Laboratory Blood Test Results
To perform the study, the research team analyzed data on blood test results for nearly 600,000 patients. Unlike most diagnostic liver algorithms, the CIRRUS model was created using a dataset comprised of patients from both primary and secondary care without the main intent of preselecting for liver disease. This renders it better suited for detecting liver disease outside a secondary care hepatology environment.
“Whilst we are all preoccupied with the coronavirus pandemic we must not lose sight of other potentially preventable causes of death and serious illness,” said Michael Moore, BM, BS, MRCP, FRCGP, Professor of Primary Health Care Research and Head of Academic Unit Primary Care and Population Sciences at University of Southampton, in the press release. Professor Moore co-authored the CIRRUS study.
“This test using routine blood test data available, gives us the opportunity to pick up serious liver disease earlier, which might prevent future emergency admission to hospital and serious ill health,” he said.
Cirrhosis (shown above in a trichrome stained micrograph) is a condition in which the liver is scarred and permanently damaged. As the condition progresses, more scar tissue replaces healthy liver tissue. This accumulated scar tissue prevents the liver from doing its primary job of regulating chemical levels in the blood and excreting bile, a substance which helps eliminate toxins from the body and breaks down fats during digestion. As cirrhosis worsens, the liver begins to fail. (Photo copyright: Wikipedia.)
According to the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), cirrhosis is most common in adults ages 45 to 54 and about 1 in 400 adults in the US live with the disease. However, the actual number may be much higher as many people are not aware they have cirrhosis, because they do not experience symptoms until the liver is badly damaged.
The NIDDK reports complications from cirrhosis include:
Portal Hypertension, a condition where scar tissue partially blocks the normal flow of blood through the liver,
“Liver cirrhosis is a silent killer. The tests used most by GPs are not picking up the right people and too many people are dying preventable deaths. We looked at half a million anonymous records and the data we needed to run CIRRUS was already there in 96% of the people who went on to have a first liver admission,” stated Sheron in the press release. “With just a small change in the way we handle this data it should be possible to intervene in time to prevent many of these unnecessary deaths.”
“Alcohol-related liver diseases are far and away the most significant cause of alcohol-specific deaths, yet currently the vast majority of people find out that their liver is diseased way too late,” said Richard Piper, PhD, Chief Executive of Alcohol Change UK, a British charity and campaign group dedicated to reducing harm caused by alcohol abuse. “What is needed is a reliable means of alerting doctors and their patients to potential liver disease as early as possible. The CIRRUS process shows real promise, and we want to see it further developed, tested and implemented, to help save hundreds of thousands, if not millions, of lives.”
CIRRUS is a true milestone in the development of computer-assisted healthcare diagnostics. It will need more research, but the University of Southampton study shows that analytical software tools can mine clinical laboratory test results that were ordered for unrelated diagnostics and identify existing health conditions that might otherwise remain hidden to the patient’s physicians.
Might this be a sign that AI platforms like Watson still cannot diagnose the wide range of patients’ conditions as accurately as a board-certified clinical pathologist?
Computer technology evolves so quickly, products often become obsolete before fulfilling their expected potential. Such, apparently, is the case with Watson, the genius artificial intelligence (AI) brainchild of International Business Machines Corp. (IBM) which was going to revolutionize how healthcare providers diagnose disease. In some areas of healthcare, such as analyzing MRIs and X-rays, AI has been a boon. But from a business perspective, Watson has failed to turn a profit for IBM, so it has to go.
In February, The Wall Street Journal (WSJ) reported that IBM is looking to sell its Watson Health unit because it is not profitable, despite bringing in $1 billion annually in revenue. The sale of Watson Health, the article states, would be aligned with IBM’s goal of streamlining the company and focusing its energies on cloud computing and other AI functions. Because one goal of the Watson project was to give physicians a tool to help them diagnose patients more accurately and faster, the problems that prevented Watson from achieving that goal should be of interest to pathologists and clinical laboratory managers, who daily are on the front lines of helping doctors diagnose the most challenging cases.
In a follow-up article, titled, “Potential IBM Watson Health Sale Puts Focus on Data Challenges,” the WSJ wrote, “… some experts found that it can be difficult to apply AI to treating complex medical conditions. Having access to data that represents patient populations broadly has been a challenge, experts told the Journal, and gaps in knowledge about complex diseases may not be fully captured in clinical databases.”
“I believe that we’re many years away from AI products that really positively impact clinical care for many patients,” Bob Kocher, Partner at Venrock, a venture-capital firm that invests in healthcare IT and related services, told the WSJ.
IBM Watson was promoted as a major resource to help improve medical care and support doctors in making more accurate diagnoses. However, in “IBM’s Retreat from Watson Highlights Broader AI Struggles in Health,” the WSJ reported that “IBM spent several billion dollars on acquisitions to build up Watson [Health] … a unit whose marquee product was supposed to help doctors diagnose and cure cancer … A decade later, reality has fallen short of that promise.”
IBM initially created Watson (above center) to be an AI tool capable of a wide variety of applications, starting with answering questions. In January 2011, Watson made headlines when it defeated previous Jeopardy champions Ken Jennings (left) and Brad Rutter (right) on the popular television game show.After its triumph, IBM announced it would transition Watson for use in medical applications and promoted it as a major resource to help improve medical care and support healthcare professionals in making more precise diagnoses. The company called its new division IBM Watson Health and stated that massive data sets would be the key to accomplish its healthcare mission. That same year, The Dark Report had an IBM executive do a presentation about Watson Health at the Executive War College on Laboratory and Pathology Management. (Photo copyright: CBS News.)
Watson’s Successes and Failures in Healthcare
During the years following Watson’s Jeopardy win, Watson Health made some positive advances in the fields of healthcare data analytics, performance measurements, clinical trial recruitment, and healthcare information technology (HIT).
However, Watson Health also experienced some high-profile failures as well. One such failure involved a collaboration with MD Anderson Cancer Center, established in 2013, to help the health systems’ oncologists develop new tools to benefit cancer patients. MD Anderson ended the relationship in 2018 after spending more than $60 million on the project, citing “multiple examples of unsafe and incorrect treatment recommendations,” made by the Watson supercomputer, Healthcare IT News reported.
Watson Health later readjusted the development and sales of its AI drug discovery tools and altered its marketing strategy amid reports of disappointing sales and skepticism surrounding machine learning for medical applications.
Underestimating the Challenge of AI in Healthcare
Since its inception, Watson Health has achieved substantial growth, mainly through a series of acquisitions. Those targeted acquisitions include:
Merge Healthcare, a healthcare imaging software company that was purchased for $1 billion in 2015,
Phytel, a health management software company that was purchased for an undisclosed amount in 2015,
Explorys, a healthcare analytics company that was purchased for an undisclosed amount in 2015, and
Truven Health Analytics, a provider of cloud-based healthcare data, analytics, and insights that was purchased for $2.6 billion in 2016.
“IBM’s Watson Health business came together as a result of several acquisitions,” said Paddy Padmanabhan, founder and CEO of Damo Consulting, a firm that provides digital transformation strategy and advisory services for healthcare organizations. “The decision to sell the business may also have to do with the performance of those units on top of the core Watson platform’s struggles,” he told Healthcare IT News.
It should be noted that these acquisitions involved companies that already had a product in the market which was generating revenue. So, the proposed sale of Watson Health includes not just the original Watson AI product, but the other businesses that IBM put into its Watson Health business division.
Padmanabhan noted that there are many challenges for AI in healthcare and that “historical data is at best a limited guide to the future when diagnosing and treating complex conditions.” He pointed to the failure with MD Anderson (in the use of Watson Health as a resource or tool for diagnosing cancer) was a setback for both IBM and the use of AI in healthcare. However, Padmanabhan is optimistic regarding the future use of AI in healthcare.
“To use an oft-quoted analogy, AI’s performance in healthcare right now is more akin to that of the hedgehog than the fox. The hedgehog can solve for one problem at a time, especially when the problem follows familiar patterns discerned in narrow datasets,” he told Healthcare IT News. “The success stories in healthcare have been in specific areas such as sepsis and readmissions. Watson’s efforts to apply AI in areas such as cancer care may have underestimated the nuances of the challenge.”
Other experts agree that IBM was overly ambitious and overreached with Watson Health and ended up over-promising and under-delivering.
“IBM’s initial approach misfired due to how the solution AI was trained and developed,” Dan Olds, Principal Analyst with Gabriel Consulting Group, told EnterpriseAI. “It didn’t conform well to how doctors work in the real world and didn’t learn from its experiences with real doctors. It was primarily learning from synthetic cases, not real-life cases.”
Was Watson Already Obsolete?
Another issue with Watson was that IBM’s marketing campaign may have exceeded the product’s design capabilities. When Watson was developed, it was built with AI and information technologies (IT) that were already outdated and behind the newest generation of those technologies, noted Tech Republic.
“There were genuine AI innovation triggers at Watson Health in natural language processing and generation, knowledge extraction and management, and similarity analytics,” Jeff Cribbs, Research Vice President at Gartner Research, told Tech Republic. “The hype got ahead of the engineering, as the hype cycle says it almost always will, and some of those struggles became apparent.”
Can Artificial Intelligence Fulfill its Potential in Healthcare?
The fact that IBM is contemplating the sale of Watson Health is another illustration of how difficult it can be to navigate the healthcare industry in the US. It is probable that someday AI could make healthcare diagnostics more accurate and reduce overall costs, however, data challenges still exist and more research and exploration will be needed for AI to fulfill its potential.
“Today’s AI systems are great in beating you at chess or Jeopardy,” Kumar Srinivas, Chief Technology Officer, Health Plans, at NTT DATA Services told Forbes. “But there are major challenges when addressing practical clinical issues that need a bit of explanation as to ‘why.’ Doctors aren’t going to defer to AI-decisions or respond clinically to a list of potential cancer cases if it’s generated from a black box.”
And perhaps that is the biggest challenge of all. For doctors to entrust their patients’ lives to a supercomputer, it better be as good as the hype. But can AI in healthcare ever accomplish that feat?
“AI can work incredibly well when it’s applied to specific use cases,” gastroenterologist Nirav R. Shah, MD, Chief Medical Officer at Sharecare, told Forbes. “With regards to cancer, we’re talking about a constellation of thousands of diseases, even if the focus is on one type of cancer. What we call ‘breast cancer,’ for example, can be caused by many different underlying genetic mutations and shouldn’t really be lumped together under one heading. AI can work well when there is uniformity and large data sets around a simple correlation or association. By having many data points around a single question, neural networks can ‘learn.’ With cancer, we’re breaking several of these principles.”
So, in deciding to divest itself of Watson Health, IBM may simply be as prescient now as it was when it first embraced the concept of AI in healthcare. The tech giant may foresee that AI will likely never replace the human mind of a trained healthcare diagnostician.
If this proves true—at least for several more years—then board-certified clinical pathologists can continue to justifiably refer to themselves as “the doctor’s doctor” because of their skills in diagnosing difficult-to-diagnose patients, and because of their knowledge of which clinical laboratory tests to order and how to interpret those test results.
