Kent Locklear, M.D.

home

authors

kent-locklear-md

Healthcare organizations (HCOs) involved in value-based care programs are trying to bend the cost curve by aggressively managing high-cost, high-need patients while eliminating waste and inefficiency without sacrificing quality.

Faced with these challenging population health management (PHM) goals, many HCOs may be considering artificial intelligence (AI) technology before it has reached a maturity level capable of delivering reliable enterprise-wise intelligence. While AI is certainly an exciting technology offering opportunities across PHM today, there is no reason to wait for its next iteration while costs and care-quality performance suffer.

Accurate, timely and reliable intelligence is available now through robust data analytics technologies that are already widely adopted across HCOs with great success. Instead of waiting for the next big thing, data analytics can meet organizations where they are now to help point providers toward improved outcomes and performance.

Overall, the key to controlling costs and improving the quality of care is to analyze data on the entire patient population. By trying to ensure that healthy patients access preventive care and patients with moderate chronic diseases are managing them, HCOs can meaningfully impact overall quality, manage their rosters and reduce the total cost of care.

An effective population health solution available now will automatically identify patient care gaps, using a combination of claims data and clinical data from multiple sources. It will also use automated messaging via phone, email or text to alert patients with care gaps that they need to make appointments to see their providers.

The combination of automated messaging and an effective use of annual wellness visits (AWVs) can motivate many people who would ordinarily not come into the office to see their providers. Aside from helping to fill care gaps, this is important because, under many value-based contracts, patients are not attributed to a provider unless they visit within a particular time period. These patients, who tend to be relatively healthy and low-cost, counterbalance the sicker patients covered under shared-savings and risk contracts.

Alerts based on analyzed data must be presented to providers very thoughtfully. Physicians are busy with patient care and prefer not to deal with data inputs they consider irrelevant. They do appreciate access to information on care and coding gaps, particularly as they apply to hierarchical condition categories crucial to Medicare Advantage plans. But to get providers’ attention, these kinds of prompts must be inserted into the clinical workflow, which requires a seamless EHR integration.

Improving outcomes across the entire patient population is aspirational and continuous, but from a practicality standpoint, HCOs may initially concentrate PHM efforts on their highest-risk, highest-need patients. Because, as most providers are well aware, just 5% of the population accounts for half of 

Data analytics can identify and categorize such populations by health risk or disease burden in a timely and useful way. Using an enterprise data warehouse, predictive analytic tools overcome the 30-day-or-more claims lag time by combining real-time, comprehensive clinical information, lab data and other information to identify patients who will generate most of the HCO’s health costs in the near term. Information on these patients can be easily shared with care-management clinicians for prompt, preventive interventions.

Furthermore, analytics delivers crucial insight into high-cost areas like emergency department utilization and post-acute-care patterns. Use is highly variable in these areas; if the population leverages these resources at above-average rates, it is important to understand why.

Physicians are being held accountable for their performance, and they’d like the opportunity to see how they’re doing on quality measures at all times. So a robust population health solution will include a dashboard that allows providers to view performance. It should also give them the flexibility to quickly drill down to the individual patient level so they can identify improvement opportunities.

Finally, neither providers nor HCOs want an array of population health solutions that don’t work together, which often is a problem with the next-big-thing solutions. Emerging technologies that only deliver risk stratification or utilization analysis, or that can’t provide actionable data, won’t cut it. Likewise, if you add too many programs — especially yet-to-be-proven solutions &mdash; to the EHR, which doctors are already spending too much time on, they won’t use them.

What physicians will use here and now is an end-to-end solution that is integrated with their EHRs and that quickly and easily gives them the information they need and provides the features necessary today to execute on population health management.

How Machine Learning Can Identify Patients at Risk of Diabetes

AI Algorithm Predicts Risk of FH with High Accuracy

Machine Learning Algorithms Predict Opioid Overdose Risk

Articles

Reliable, actionable data analytics can get the job done.

Kent Locklear, M.D.

Articles

Why Wait for the Next Big Thing? Improve PHM Now.