Driving value across the enterprise: Health execs have a laser focus on analytics

My Take

Driving value across the enterprise: Health execs have a laser focus on analytics

By Steve Burrill, vice chairman, US health care leader, Deloitte LLP

The ability to tap into data analytics has long held the promise of guiding hospitals and health systems toward clinical, operational, and financial efficiencies. Hospitals and health systems are generating mountains of data, and data-driven decision-making could help to transform their businesses and the care they provide to patients.

The term “analytics” refers to the systematic use of technologies, methods, and data to derive insights and to enable fact-based decision-making for planning, management, operations, quality, safety, measurement, and learning. Globally, the health care analytics market has tripled over the past three years—from between $4 billion and $5 billion in 2015 to about $15 billion in 2018.¹ The market is expected to top $50 billion by 2024.²

But taking advantage of data analytics goes beyond investments, developing a corporate vision, and buying the appropriate software. Health plans and health systems also need expertise. Late last year, the Deloitte Center for Health Solutions surveyed 56 health system chief information officers (CIOs), chief technology officers (CTOs), and chief analytics executives to understand the importance of analytics and where organizations are focusing their efforts. Nearly one-third of our respondents said they had a C-Suite position dedicated to analytics—up from just 12 percent in 2015. Over the next three years, 29 percent of health system executives expect to hire data scientists, and 21 percent intend to bring on visualization developers and/or data architects.

Most health system execs see analytics as the future

According to our survey results, 36 percent of health system executives consider analytics to be an important part of their organization’s strategy, but the majority (84 percent) expect it will be essential three years from now.

We also found that among health systems:

• 70 percent have a defined vision for data analytics (vs. 48 percent in 2015)
• 88 percent have dedicated analytics departments (vs. 79 percent in 2015)
• 68 percent have centralized governance models (vs. 59 percent in 2015)

While interest in analytics and adoption of supporting technology has clearly accelerated over the past three years, some health systems are further ahead than others. Although 84 percent of executives told us that they think analytics will be extremely important in the next three years, about 30 percent of organizations do not have an integrated analytics strategy. This could be a missed opportunity. According to a separate survey of hospital and health system leaders conducted by HIMSS, 57 percent of executives said they are using predictive analytics, and they expect the technology could reduce organizational costs by at least 15 percent over the next five years.

Health plans see analytics as a differentiator 

Similar to health system executives, a majority of health plan leaders (67 percent) said data analytics is extremely important to their organization as a competitive differentiator, according to a 2017 Deloitte survey. A health plan with a large Medicare Advantage (MA) book of business attributes its consistently high performance on Star ratings to its analytics. Dedicated analysts keep constant watch on data that drive specific Star measures. If they see a downward trend, or if the tolerance threshold for that measure is crossed, the analysts raise the issue with the internal operations team (such as medical management) that can intervene and correct the trend. Information gleaned from claims data can lead to increased efficiencies, improved affordability/reduced medical costs, or enhanced customer engagement/experience, they said. About 75 percent of our health plan respondents expected their analytics spending would increase in the year ahead.

Hospitals and health systems that are lagging their competitors should consider developing a strategy to incorporate analytics. Health systems that already have mature analytics capabilities could move further ahead by incorporating artificial intelligence (AI) and predictive analytics. Here’s a look at how analytics are being used:

• Providence St. Joseph Health (PSJH), which operates 51 hospitals across seven states, has been building analytics capabilities over the past several years. It is transitioning its entire system to one EHR platform and is investing in a data lake for all data sources. The health system is piloting a program that integrates the system’s supply chain database with its EHR to support a pop-up for physicians showing the wholesale cost of an ordered drug or test, along with pricing for comparable treatments. The pilot showed that physicians choose the lower-cost option when efficacy is comparable, resulting in a marked reduction in cost of care.
• Hartford HealthCare (HHC), a six-hospital integrated delivery system in Connecticut, had gone through some of the growing pains typically associated with multiple acquisitions and disparate software systems. HHC leaders determined that an analytics platform could generate data that might help achieve goals related to value-based care, patient experience, growth, quality, safety, and efficiency. Its focus and vision for the past few years has been getting information into the hands of decision-makers for both clinical and financial performance. The health system is building its machine-learning capabilities and is leveraging its standardized data sets to improve patient access and quality of care. In the next three to five years, HHC leaders hope to have both machine learning and AI capabilities in place.

Hospitals and health systems are increasing their use of data sources and are expanding the types of data sets they use for analytics. According to our research, in 2018, the top three external data sources that were integrated into analytics were: health plan claims (73 percent), US Centers for Medicare and Medicaid Services (CMS) data (73 percent), and financial benchmarking data (71 percent). The use of social determinants of health (SDoH) data is expected to grow most in the next three years: While 38 percent of respondents said their organizations use it currently, 77 percent anticipate they will use it in three years.

Analytics is no longer relegated to the IT department. As the health care sector continues its march to a value-based model, the ability to tap into data analytics will likely become even more critical. During the past three years, we have seen bigger investments in analytics infrastructure, resources (dollars and people), and scope. Over the next three years, the majority of our survey respondents expect analytics to become even more critical.

Email | LinkedIn

¹ IQ4I Research & Consultancy, Healthcare analytics global market, 2014; MarketsandMarkets, Healthcare analytics market by type (predictive, prescriptive, cognitive), application (clinical, RCM, claim, fraud, waste, supply chain, PHM), component (service, software), delivery (on-demand, cloud), end user (payer, hospital): Global forecast to 2024, accessed March 12, 2019; ResearchFox, “Healthcare analytics market outlook,” February 2015.

² Jeff Lagasse, “Healthcare business and financial analytics market to hit $50B by 2024,” Healthcare Finance, accessed March 12, 2019.

Back to top

In the News

CMS boosts 2020 Medicare Advantage payments by 2.5 percent

On April 1, CMS released the final 2020 Rate Announcement and Call Letter for Medicare Advantage (MA) and Medicare Part D. On average, MA plans will see a 2.53 percent increase in their payments for the 2020 plan year. This is higher than the increase listed in the earlier notice, due primarily to a higher rate of growth in fee-for-service spending.

The agency also finalized several policies, including the one giving MA plans the ability to provide more tailored services to chronically ill beneficiaries. The notice also addresses aspects of the opioid epidemic. CMS is encouraging MA plans to offer supplementary benefits and cost-sharing reductions aimed at patients who have chronic pain or who are undergoing addiction treatment. The agency is also calling on Part D plans to reduce cost-sharing for opioid-reversal agents. These changes were announced in Part II of the Advance Call Letter in January (see the February 5, 2019 Health Care Current).

As proposed in Part I of the Advance Call Letter, CMS will give equal weight to diagnoses coming from encounter data and fee-for-service data (the traditional source of information used in the risk-adjustment model) to determine patients’ risk scores beginning in 2020. Also, as required by the 21st Century Cures Act, CMS finalized a methodology that accounts for the number of medical conditions a patient has in the risk model (see the January 8, 2019 Health Care Current).

(Source: CMS, CMS finalizes Medicare Advantage and Part D payment and policy updates to maximize competition and coverage, April 1, 2019)

Back to top

Telehealth gained traction in 2017, according to new report

Growth in telehealth utilization outpaced all other care venues between 2016 and 2017, according to a new report from FAIR Health, a national non-profit health information organization. According to the report, the use of telehealth grew by nearly 53 percent, and private insurance claims for those services grew nearly 1,200 percent, between 2016 and 2017. Utilization of urgent care centers, retail clinics, and ambulatory surgery centers (ASCs) increased by 14, 7, and 6 percent, respectively. Emergency department (ED) use decreased by 2 percent during the same time.

Telehealth can be a convenient way to access certain health care services. In 2017, use of telehealth services grew nearly two times faster in urban areas, compared to rural areas. People were using telehealth services for different diagnoses in 2017 compared with 2016, the report found. In 2017, injuries, acute respiratory infections, and digestive issues were among the top telehealth diagnostic categories, each accounting for 13 percent of diagnoses. Behavioral health issues, which was the top diagnostic category in 2016, dropped to fifth in 2017, accounting for 7 percent of diagnoses.

(Source: FAIR Health Inc., “FH Healthcare Indicators® and FH Medical Price Index® 2019,” accessed April 2019)

Back to top

HRSA publishes ceiling prices for 340B drugs

On April 1, the Health Resources and Services Administration (HRSA), an agency in the US Department of Health and Human Services (HHS), updated a website that health care facilities can use to register for the drug-discount program. The Office of Pharmacy Affairs 340B Information System now allows 340B hospitals and other providers to confirm the accuracy of drug manufacturers’ ceiling prices. According to the American Hospital Association (AHA), the website will help prevent drug manufacturers from overcharging hospitals for 340B drugs. In January, HHS finalized a rule setting ceiling prices for 340B, which required HRSA to update the website (see the February 26, 2019 Health Care Current). The 340B program gives certain hospitals and clinics access to discounts from drug manufacturers. In 2010, the Affordable Care Act (ACA) broadened the definition of covered entities that could participate.

Back to top

House committee holds hearing to discuss ‘surprise billing’

The House Education and Labor Committee held an April 2 hearing on surprise medical bills, during which both Democratic and Republican lawmakers discussed legislation. Surprise medical bills can occur when an out-of-network physician provides care at a rate not negotiated by the patient’s health plan. This seems especially common during emergency visits. During the hearing, lawmakers discussed the merits of setting payment rates or implementing a “baseball-style” arbitration method for addressing out-of-network services—each side proposes a price and an arbiter makes a choice that binds both parties. Several states have implemented baseball-style arbitration for pricing out-of-network bills.

Lawmakers also discussed implementing a single-payment rate or using a bundled-payment model with one up-front fee for a set of medical services. The same day as the hearing, the AHA, American Medical Association (AMA), and Federation of American Hospitals (FAH) sent a letter to committee leadership opposing bundled-billing. According to the letter, hospital bundled-billing practices could be administratively complex, while providing insufficient protection to patients.

In February, a bipartisan group of senators sent letters to health plans and hospitals to request data on surprise billing practices. That same month, the president discussed efforts targeting surprise billing in his 2019 State of the Union address (see the February 12, 2019 Health Care Current). Several stakeholder groups have proposed legislation preventing this practice, such as passing legislation banning balance billing—where patients are charged beyond their cost-sharing obligations—and increasing price transparency to patients regarding potential out-of-network care (see the February 26, 2019 Health Care Current).

Back to top

CMS approves more partial Medicaid expansion, work requirements

On March 28, CMS approved Utah’s application to partially expand its Medicaid program to adults with incomes up to 100 percent of the federal poverty level (FPL). The expansion went into effect on April 1 (see the February 19, 2019 Health Care Current). Utah’s application allows the state to cap enrollment for the expansion population if the state lacks enough funding to match federal payments. Utah is the first state approved for this provision. Additionally, the state will receive only the standard federal-match rate of 70 percent for the partial-expansion population, rather than the 94 percent rate called for by the ACA for full Medicaid expansion (adults up to 138 percent FPL). CMS’s approval arrived a day after a US District Court judge struck down work/community engagement requirements in Arkansas and Kentucky (see the April 2, 2019 Health Care Current).

Utah’s waiver also includes a work requirement. Beginning on January 1, 2020, people who qualify for coverage under Utah’s expanded eligibility rule must fill out an online form that directs them to a training program. Beneficiaries must complete 48 job searches during their first 90 days of Medicaid eligibility. Unlike the other states that have received CMS approval to implement work requirements, Utah enrollees will not have to report a minimum number of hours of monthly employment (see the March 12, 2019 My Take).

Back to top

House Committee passes bipartisan bills to lower drug prices, drive competition

On April 3, the House Energy and Commerce Committee reached a bipartisan agreement to advance legislation aimed at reducing prescription drug prices by increasing competition from cheaper generic drugs. Committee Chairman Frank Pallone (D-NJ) and Ranking Member Greg Walden (R-Ore.) will allow six bills to advance to a full committee vote.

One of the bills, the Creating and Restoring Equal Access to Equivalent Samples (CREATES) Act, targets drug manufacturers that misuse safety guidelines or refuse to provide samples of their products to generic drug makers to hinder development of less-costly alternative therapies (see the March 19, 2019 Health Care Current). The CREATES Act was reintroduced in the House and Senate in February. During a Senate Finance Committee hearing held the same month, a panel of seven witnesses—all of whom are executives at large pharmaceutical manufacturing companies—expressed support for the bill (see the March 5, 2019 Health Care Current).

Another bill, the Protecting Consumer Access to Generic Drugs Act, prevents brand-name drug manufacturers from paying generic drug makers to keep their generic equivalents out of the market, a practice known as “pay-for-delay.” House Democrats and Republicans reached an agreement to ban future settlements. The bill initially made pay-for-delay agreements as far back as 2013 illegal. The committee also approved changes to the Orange Book Transparency Act to allow patents for drug-delivery devices to remain in the Orange Book, which lists patents for approved drugs.

The other bills are:

• H.R. 1781, the Payment Commission Data Act of 2019, which allows the Medicare Payment Advisory Commission (MedPAC) and the Medicaid and CHIP Payment and Access Commission (MACPAC) to access drug-rebate data for research.
• H.R. 938, the Bringing Low-cost Options and Competition while Keeping Incentives for New Generics Act of 2019, discourages generic first filers from “parking” applications and delaying the start of their 180-day generic exclusivity (see the March 19, 2019 Health Care Current).
• H.R. 1520, the Purple Book Continuity Act of 2019, codifies publication of biopharmaceutical patents in the Purple Book in a format similar to that of the Orange Book.

Back to top

Breaking Boundaries

Voice analysis can provide clues to health

An Israeli company called VoiceSense is using real-time voice analysis to evaluate and predict human behaviors—and identify health risks. VoiceSense and similar companies, along with researchers, are taking advantage of the plethora of devices that capture human speech, such as our mobile phones and the digital assistants many people have in their homes.

Voice is personal and difficult to fake. The key to voice-analysis research is capturing individual tones, speed, emphases, and pauses, and applying machine learning to make predictions. Researchers are looking at people who have anxiety, as well as people without anxiety, and feeding the data to an algorithm. Over time, the algorithm learns to pick up the subtle speaking signs that might indicate whether someone is anxious. Using voice to identify anxiety, depression, and specific conditions such as Parkinson’s disease or post-traumatic stress disorder (PTSD) is still in the early stages.

Mental and behavioral health issues can be hard to monitor. Signs and symptoms can creep up on people who have depression before they realize they might need help. Having a sensor that could monitor and alert a person to such issues could enable earlier intervention. A few startups, such as Sonde Health and Ellipsis Health, are designing mental-health monitoring systems that use voice and other factors to track signals such as mood, diminished interest, avoidance, and fatigue. The information is sent to the patient and the physician. Some companies are starting to publish early results in peer-reviewed journals.

Some companies and researchers are also pursuing other non-health-related uses for voice analytics, such as determining the likelihood of someone defaulting on a bank loan. The concern with such uses is that machine learning and algorithms are not yet perfect, and predictions can be wrong. A next step for developers of this technology is to validate the tools in different populations and for different uses. Some stakeholders are concerned about privacy violations and discrimination, as voice is considered a biometric measure. Some states have laws in place that offer protections for biometric privacy and security.

(Source: Angela Chen, “Why companies want to mine the secrets in your voice,” The Verge, March 14, 2019)

Back to top