When Amazon envisioned Alexa, an AI-powered, voice-activated customer recommendation system, it was a feat that required machine learning and massive amounts of data to provide answers to conversational queries quickly, even in a noisy environment. Now, the same data analysis capabilities that enable Amazon to become hyper-familiar with consumer purchasing patterns could hold the key to reducing waste in healthcare.
Think about the similarities between healthcare and retail. Both industries revolve around the consumer, and they use data to gain context into behavior and draw meaningful conclusions. In healthcare, this includes the ability to predict which consumers could develop type 2 diabetes with 95% accuracy or to pinpoint where and when the Covid-19 virus will spread and how to protect those most vulnerable.
Reducing Clinical and Staff Burnout with AI Automation
As technology advances, AI-powered tools will increasingly reduce the administrative burdens on healthcare providers.
The same intelligence that empowers E-retailers to tap into knowledge of consumer purchasing patterns to make complementary product recommendations can empower health plans and healthcare providers to:
- Direct members to healthcare organizations that demonstrate high quality at lower costs
- Pair patients with physicians who specialize in their condition, increasing efficiency while reducing expense
- Understand where waste originates and develop guardrails that help curtail excess spending
- Uncover suspected fraud, such as when laboratory tests are upcoded or higher-than-necessary quantities are ordered to receive higher reimbursement
But while providers and health plans have a tremendous amount of data at their disposal—from claims data to demographic data to information around social determinants of health—it takes more than an AI software solution to gain meaningful insight. At a time when 25% of healthcare spending can be considered wasteful, providers and health plans should consider these three approaches to optimize the use of machine learning to boost value.
Cast a wide net
Make sure that the data being aggregated covers a wide universe. Ideally, data sources should range from critical healthcare transactions to patient wearables, demographic data and social determinants of health.
The Funding Model for Cancer Innovation is Broken — We Can Fix It
Closing cancer health equity gaps require medical breakthroughs made possible by new funding approaches.
Attention to data aggregation is especially important during the pandemic when the types of anomalies that hospitals and health plans might uncover likely are very different from before the pandemic simply because the types of care being provided have changed. For example, the accelerated shift toward virtual care delivery led to a 2,980% increase in telehealth claims during the pandemic. This intensifies the need for machine learning tools that can spot “telefraud” schemes—which rose substantially during the pandemic, particularly among Medicare and Medicaid beneficiaries. It also necessitates access to a broad range of data that can be leveraged by machine learning software in detecting suspicious patterns that can point to fraud, waste or abuse in an evolving environment.
Continually refresh the data
For machine learning to make relevant, timely inferences, the data must be refreshed constantly. Otherwise, it rapidly starts to become stale, limiting the potential for meaningful analysis that reduces waste. That’s especially important during Covid-19, when usage patterns have changed leading to a greater emphasis on virtual care, a significant drop in elective procedures and new types of treatments related to the coronavirus. The more data is available for analysis, the better able machine learning applications become in detecting patterns that can suggest waste or even fraud or abuse.
For example, some health plans use machine learning to spot spikes in Covid-19 claims by facility. Machine learning also has proved valuable in finding instances where a provider has billed for services outside the provider’s specialty area of expertise, warranting closer review. When suspicious claims are identified, machine learning software flags claims for internal review by health plan nurses or coders.
Look for ways to make data insight accessible to providers at the point of care
Machine learning gives healthcare organizations and health plans enhanced ability to predict Covid-19 health risks, from a specific population’s risk of contracting the disease to the health outcomes that might be observed. Just as E-retailers broaden their views of consumer data beyond individual purchasing patterns to seek evidence of larger trends, machine learning could empower providers and health plans to determine ways to reduce risk for specific populations. This could mean driving faster access to care for consumers with specific comorbidities or providing clinical decision support that could help improve outcomes. It can also include engaging consumers when the data captured by remote monitoring devices suggests that they are in the early stages of disease, such as heart failure or arrythmias. These approaches help direct interventions to vulnerable consumers sooner, reducing overall cost of care.
At the University of California, Irvine, researchers developed a machine learning model to help predict the likelihood that a Covid-19 patient’s condition will worsen to the point of ICU admission, ventilation or death. This model also alerts providers to factors that are most associated with severe Covid-19 cases, including advanced age, gender (men are more likely to develop severe symptoms), hypertension, diabetes and coronary artery disease.
Reducing Waste, One Consumer at a Time
Just as Amazon leverages machine learning to develop valuable insights into consumer purchasing behavior in real time, healthcare organizations can strengthen their ability to reduce costs of care and eliminate waste and abuse with this type of intelligence. Today, machine learning’s value proposition lies not just in the ability to remove excess cost from the system, but also in improving the ability to provide the right care at the right time for vulnerable populations. For small and large organizations alike, democratization of this technology will be key to optimizing care and spending during the COVID-19 pandemic and beyond.
Photo: Flickr, Cerillion Skyline
Nicole Neumarker is executive vice president, development and innovation, Cotiviti.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.