Taking the robot out of the human: Meet the health care workforce of the future | Deloitte US has been added to your bookmarks.
Taking the robot out of the human: Meet the health care workforce of the future
Health Care Current | May 1, 2018
This weekly series explores breaking news and developments in the US health care industry, examines key issues facing life sciences and health care companies and provides updates and insights on policy, regulatory, and legislative changes.
Taking the robot out of the human: Meet the health care workforce of the future
By Sarah Thomas, Managing Director, Deloitte Center for Health Solutions, Deloitte Services LP
Based on some of the headlines I’ve come across recently, it seems we are all about to be replaced by robots and artificial intelligence (AI). From NBC News, “Will robots take your job? Humans ignore the coming AI revolution at their peril.”1 Or this one from The New York Times, “Will robots take our children’s jobs?”2
These stories paint a pretty dire picture of the future workforce. But the future for health care employees might be much more optimistic. Rather than replacing human workers, technology could help health care organizations create a more professionally satisfied workforce—by taking the robot out of the human.
Labor is a major part of health care costs. Among hospitals, labor is typically the largest line item in any hospital’s budget, and accounts for almost 60 percent of noncapital costs,3 on average, and costs are growing. Health plans and life sciences companies also employ many people, with a variety of jobs. Some of these are highly focused on complex and innovative work, but many of these have aspects that are routine and repetitive.
Health care jobs are evolving, and they are likely to change even more as technology removes some of these routine and repetitive tasks. We recently held a Dbriefs webcast to explore how emerging technologies could impact the workforce at hospitals, health plans, and life sciences companies.
The term “no-collar workforce” is used to describe a strategic combination of humans and machines that results in enhanced performance and drives greater efficiencies. Another way to think about this is how to take the robotic (routine, repetitive) functions out of the jobs now performed by humans.
What kinds of technologies are we talking about?
The spectrum in the chart above starts with robotic process automation (RPA), which is the easiest cognitive technology to implement. It is effective for high-volume, repetitive, rules-based tasks. Users can see a relatively quick payback in the form of reduced costs, which is an important goal for many organizations.
Moving along this spectrum, there are increasing levels of intelligent cognitive automation, which can help handle less structured data and more complex activities. Natural language processing is where a machine learns to read and pull relevant information from documents. This technology could help streamline processes related to contracts with providers, employers, health plans, or suppliers. Software, for example, can extract information from very lengthy documents to develop insights.
Machine learning is used when there are large amounts of data, and the machine learns the best way to process it. This technology also begins to understand various data formats. It also can learn from humans about how to respond to different scenarios. Finally, artificial intelligence is the ability of machines to behave in a way that imitates intelligent human behavior.
Robots could make us less robotic
Here is an overview of how these technologies could impact the workforce of the future for health care and life sciences companies:
Health plans: These organizations are adopting technologies across the entire automation spectrum—typically starting in the shared-service areas. For RPA, adoption is happening in finance areas that have many high-volume repetitive activities, such as monthly journal entries to update the general ledger. Another area for automation is in human resources operations, which might include managing the leave-of-absence process for employees.
Often, health plans will use these technologies for back-office functions, such as enrollments, billing, and claims processing. The goal is to reap the benefits of core operational areas by automating those processes and freeing employees from mundane tasks such as data entry and reconciliation. Staff members should then have time to conduct insightful analyses. They also can focus on improving the customer experience by spending more time taking care of members. For many employees, this is what they really want to be doing.
We also are seeing some activity around machine learning. This technology can help health plan leaders understand how their employees and customers are interacting with the organization. Machine learning could, for example, anticipate a member’s needs and offer suggestions to customer service representatives as they are talking to members, or even to providers who are interacting with the health insurer. Machine learning could also allow brokers and sales agents to offer real-time quotes when interacting with potential customers.
Health systems: Some health care providers are beginning to take advantage of technology to transform their workforce in big and small ways. Like health plans, providers are using RPA for functions such as revenue cycle, but we also are starting to see it used in care delivery.
While about 90 percent of registered nurses say they are satisfied with their career choice, one out of three are unhappy with their current job, according to a survey of 3,400 nurses conducted by staffing firm AMN Health Care.
Clinicians typically spend about a quarter of their work day completing tasks that are not critical to patient care. That means they often can’t respond quickly to call buttons. Using artificial intelligence, Deloitte worked with clients to revolutionize the nurse call button. Patients can speak to the device and explain what they need (e.g., medication, help getting to the restroom). The patient receives an immediate response to the request, so they know that it has been heard. Critical details are smart-routed to the care team, including the priority and urgency of the requests.
This smart call button can change the way the clinicians’ work is processed, and it reduces or even eliminates some tasks. This can give clinicians more time to spend caring for patients, which should help improve patient experiences and outcomes.
Life sciences companies: Finance is certainly an area where we are seeing adoption, but life sciences companies are also building Centers of Excellence and are looking at use cases across the spectrum, including supply chain, research and development (R&D), and commercial. Examples in R&D include protocol development, the cognitive analytics to comb through all published and proprietary data to build a protocol for clinical trials, as discussed in Deloitte’s recent research. Life sciences companies can use natural language processing to build clinical study reports.
In the manufacturing supply chain area, we see use cases for emerging technologies in quality control and assurance. Some medical device companies, for example, are looking into how augmented reality could be used to train people to use their products.
On the commercial side of the business, applications of these technologies could be used to set reimbursements for 340B validation. Cognitive automation could help avoid ineligible payments.
Can technology improve job satisfaction?
I think all of these examples are exciting. While we are not seeing wholesale changes to jobs yet, removing routine and repetitive duties should help to make people happier in their professional lives. I predict that the use cases and adoption of these technologies will accelerate over time, and will eventually change not just what people do in their jobs, but also the types of careers people will have in life sciences and health care.
As Maureen Medlock, one of our speakers on the webcast, said, “I would challenge leaders to shift their mindset. Instead of focusing on questions like, ‘how can we hire and retain more employees to combat labor shortages?’ [they should] instead ask, ‘how can we automate and innovate so that high-performing people would not want to work anywhere else?’”
Despite recent articles, I don’t expect robots are going to push us all out of the workforce, but I do think emerging technology could help remove some of the more tedious elements of our jobs. I hope that the adoption of technology will lead to happier health care clinicians and administrators, happier and healthier patients, and less expensive and more efficient care.
In the news
CMS Innovation Center issues RFI on direct provider contracting models
The US Centers for Medicare and Medicaid Services (CMS) is seeking stakeholder feedback on direct provider contracting (DPC) between health plans and primary care or multi-specialty group practices. To that end, the agency issued a Request for Information (RFI) on April 23. Ultimately, CMS might use the feedback as it tests a DPC approach within traditional Medicare, Medicare Advantage (MA), and Medicaid.
Under a DPC model, CMS might pay a fixed monthly fee to practices based on the number of beneficiaries. The agency could contract directly with primary care or multi-specialty group practices that agree to be accountable for the cost and quality of care for a specific population. Unlike existing primary care models, a DPC approach would let practices to take on “two-sided” financial risk.
In its RFI, CMS says such models could improve the revenue stream for practices by giving them more flexibility in determining where and how they provide care. DPC, as outlined in the RFI, aims to support the shift to value-based payment models while reducing physicians’ administrative burdens. Some stakeholders, including consumer organizations, have cited concerns. In particular, some have raised the possibility that beneficiaries might have to pay more to see clinicians under this model.
Comments are due on May 25. The RFI asks for input in six areas:
(Source: Center for Medicare and Medicaid Innovation, Request for Information on Direct Provider Contracting Models, April 23, 2018)
CMS proposes series of overhauls to Prospective Payment Systems, quality provisions
CMS last week proposed revisions to its payment policies and rates under the Inpatient Prospective Payment System (IPPS) and Long-Term Care Hospital Prospective Payment System (LTCH PPS) for FY 2019.
The proposed rule would raise IPPS payments by 3.4 percent, increasing Medicare spending by $4 billion. It would also raise the LTCH PPS payment rate by 1.15 percent. However, CMS noted that LTCH PPS payments would decrease slightly overall (0.1 percent, or $5 million) in FY 2019 due to other changes in the rule called for in the recent budget bill.
Under the proposed rule, CMS would increase uncompensated care payments to $8.25 billion in FY 2019, a $1.5 billion increase from FY 2018, due to an updated forecast and new data on the uninsured rate.
Increasing price transparency is also an aim of the proposed rule. It requires hospitals to release a standard list of prices for their medical services, which can vary widely across the US and even among the same geographic area.
As part of the rule, CMS included a Request for Information to obtain recommendations for interoperability (sharing health care data among providers) using hospitals’ Conditions of Participation. The proposed rule would make changes to the Electronic Health Record Incentive program to promote interoperability and make the incentive program more flexible and less burdensome by placing a strong emphasis on measures that require the exchange of health information between providers and patients. The proposal would rename the program “Promoting Interoperability” to reflect this revision.
The proposed rule would also eliminate a number of quality reporting measures that the agency concluded were duplicative, overly burdensome, or “topped out” (meaning nearly all providers receive high scores).
Instituted as part of the Affordable Care Act, the requirement can also be fulfilled if hospitals allow the public access to the data after an inquiry, CMS said.
The agency stressed that aspects of the rule would be part of a broader goal to promote value-based care.
(Source: CMS, 42 CFR Parts 412, 413, 424, and 495)
FDA sees future in digital health and AI, Gottlieb says
The US Food and Drug Administration (FDA) is taking steps to support the use of digital health tools and artificial intelligence (AI) in the development of drugs and medical products. FDA Commissioner Scott Gottlieb, M.D. announced some of these steps on April 26 at AcademyHealth’s annual Health Datapalooza conference in Washington, D.C. He told attendees that the agency wants to give companies and innovators the regulatory predictability they need to make long-term investments and other risks that support innovation. Digital health, he added, offers opportunities to improve medical outcomes, enhance efficiency, and reduce costs. The agency intends to streamline the regulatory path for digital health products.
Gottlieb said that AI holds enormous potential for the future of medicine, and that the agency is developing a regulatory framework to support the use of AI-based technologies. In February, the FDA approved a type of clinical decision support software that uses AI algorithms to notify a neurovascular specialist faster than would be possible without the software, potentially decreasing time to treatment.
RELATED: On Friday, the FDA announced two new components of its Digital Health Innovation Action Plan. One component is the release of a progress update on the software precertification pilot program, the agency’s proposed voluntary pathway for pre-certifying companies to allow for a more streamlined review of software as a medical device. The agency is also asking for public comment on this model. The second component is draft guidance on multiple function device products. Medical products may contain several functions, some of which fall under the FDA’s regulatory oversight as medical devices, while others, in isolation, are not subject to this oversight. This draft guidance offers additional clarity to the digital health community about where the FDA sees its role in digital health, and where the agency will not be involved.
(Source: FDA, Transforming FDA’s Approach to Digital Health, Remarks by Scott Gottlieb, M.D., AcademyHealth’s 2018 Health Datapalooza, Washington, DC, April 26, 2018)
CMS releases Medicare Advantage data to researchers
CMS Administrator Seema Verma spoke at last week’s AcademyHealth Datapalooza conference. Her remarks focused on the agency’s release of Medicare Advantage encounter data to researchers to help increase our understanding of health care trends among seniors. Since 2012, CMS has been collecting encounter data (similar to claims) on patient conditions and services to help calculate payments for MA health plans. However, this is the first time the agency has released the data to researchers.
CMS has released preliminary MA encounter data from 2015. The agency expects to release additional heath plan data through August, with final reports to follow. CMS plans to release MA data annually to researchers who have signed a data-user agreement. Next year, CMS intends to make Medicaid and Children's Health Insurance Program (CHIP) data available, according to Verma.
CMS follows up on critical GAO evaluation of 1115 waivers
CMS is improving the way it evaluates §1115 Medicaid demonstrations, an agency representative said at the Medicaid and CHIP Payment and Access Commission’s (MACPAC) most recent meeting.
The session came in response to a February analysis by the Government Accountability Office (GAO) that said CMS does not evaluate whether Medicaid demonstrations reduce spending or improve the quality of care sufficiently (see the February 27, 2018 Health Care Current).
The CMS representative presented the agency’s response to each recommendation from the GAO report. Those responses are outlined below:
CMS provides states with technical assistance for their evaluations. The agency is also developing guidance that states can use in their evaluations, which it will release online over the next year. It also is developing ways to disseminate evaluation results among partners and other agencies.
Under §1115 of the Social Security Act, states may request the federal government’s permission to waive certain requirements for Medicaid and implement demonstrations that seek to fulfill the program’s goals more effectively.
Related: A 2017 Deloitte paper discusses how Medicaid alternative payment models (APMs) are improving access and care.
HHS releases second installment of state funding to combat the opioid crisis
On April 18, the Department of Health and Human Services (HHS) released $485 million in grants to help 50 states and six US territories continue the battle against the opioid crisis.
The Opioid State Targeted Response (STR) grants were created by the 21st Century Cures Act. The grants are administered by the Substance Abuse and Mental Health Services Administration (SAMHSA) within HHS. During the first round of funding, recipients used the money to support evidence-based prevention and treatment programs and practices. Grantees also used funds to implement effective medication-assisted treatment, to promote the use of naloxone and key prevention strategies, and to build sustainable systems of recovery support services across the country.
RELATED: Opioid misuse legislation moves forward
Last week, both the House and Senate passed separate bills targeting the opioid crisis. The Senate Health, Education, Labor and Pensions Committee (HELP) passed its package on April 24. The next day, the House Energy and Commerce Health Subcommittee approved 57 measures. The bills propose changes to the way federal agencies regulate, oversee, and enforce rules on opioid painkillers. They also address the ways in which states try to mitigate abuse as well as improve Medicaid coverage for patients with opioid-use disorder.
States address racial disparities in maternal death rates
Serena Williams’ near-death experience after the recent birth of her daughter is a reminder that childbirth-related morbidity rates are high among black women. The professional tennis player experienced a pulmonary embolism in February—shortly after leaving the hospital with her new baby.
Nationwide, black women are more than three times likelier to die from causes related to pregnancy or childbirth than are other women. According to federal data, 42.8 black women die for every 100,000 successful births, compared with a rate of 12.5 for white women, and 17.3 for women of all other races.
Maternity health care disparities, regardless of social and economic status, are on the rise. Studies show college-educated black women still had higher rates of life-threatening conditions during delivery than women of other races who never graduated high school.
Some states are taking action to address these disparities. In New Jersey, the Department of Health launched the “Improving Pregnancy Outcomes Initiative” five years ago to help local providers connect women with high-risk pregnancies to appropriate care. The program is now active in 21 counties. Health officials in New York are piloting a Medicaid initiative that will cover doulas (birth coaches who provide physical and mental health services), the state announced last week. Minnesota and Oregon already cover doula services in their Medicaid programs. One challenge is making women and clinicians aware of these services.
Two federal bills—that would help states create maternal task forces and look at gaps in care and treatment—are stalled in committee. However, at least five states have pending legislation to address this issue. Last month, Indiana Governor Eric Holcomb (R) signed a bill to create a maternal mortality review committee to scrutinize deaths and near-deaths among expectant and new mothers. The committee also is expected to make policy recommendations to help improve maternal health. About 35 states have established review committees or are in the process of doing so. In Michigan, a local nonprofit called Make Your Date Detroit is committed to helping women have healthy pregnancies and births. The organization runs prenatal education classes, arranges free rides to medical appointments, and connects expectant mothers with mentors who can help them through each phase of the pregnancy.
RELATED: Medicare's Hospital Readmissions Reduction Program, which financially penalizes hospitals that have relatively high readmission rates, has helped to reduce racial disparities in hospital readmissions, according to a study published in the April issue of Health Affairs. The research team examined trends in 30-day readmission rates for congestive heart failure, acute myocardial infarction, and pneumonia among non-Hispanic white and non-Hispanic black patients, and among hospitals that serve large populations of minorities. Looking at the different groups, researchers found lower readmission rates for both black and white patients. Readmissions fell, on average, 0.45 percent per quarter for black patients, and 0.36 percent for white patients during the penalty-free implementation period of April 2010 to September 2012, compared to pre-implementation of the program.
Evidence shows that before the program, black patients had, on average, 20 percent higher readmission rates than white patients. Hospitals serving a higher proportion of black patients also had higher readmission rates than other hospitals—even after controlling for patient complexity.
Minority-serving hospitals made more improvements than other hospitals during the program. But, they were still more likely to be penalized because the program rewards hospitals based on their ranking relative to each other and not based on their own improvement over time, according to the study. The authors speculated that minority-serving hospitals' lack of resources may hinder their efforts to reduce readmissions.
Previous studies have demonstrated that the Hospital Readmissions Reduction Program is linked to improved readmission rates, yet prior to this study little was known about its effect on racial disparities. The results suggest that more work needs to be done to ensure that pay-for-performance programs improve equity in care.
(Source: José F. Figueroa, Jie Zheng, E. John Orav, Arnold M. Epstein, Ashish K. Jha, “Medicare program associated with narrowing hospital readmission disparities between black and white patients,” Health Affairs, April 2018; Nina Martin and Robin Fields, Here’s one issue blue and red states agree on: Preventing deaths of expectant and new mothers, March 26, 2018)