Managing health for millions

In the 1980s and ‘90s, managed care companies first sought to use “disease management” strategies as a way to deliver better healthcare at lower costs. But these early approaches never caught on with customers for two reasons. First, they failed to meaningfully “engage” the customers in the process and second, their “disease only” focus missed the fact that people struggling with serious or chronic conditions like obesity, heart disease, diabetes, asthma, or cancer often experience anxiety, depression, or other behavioral health issues that can interfere with adherence to medical treatment.

Quadir “JJ” Farook, president and CEO of InfoMC (Conshohocken, Pa.), a developer and provider of payer-focused health analytics and disease management software, explains, “If you had cancer, you would have a cancer disease manager from one company. If you were diabetic, you would have a diabetes disease manager from another. You had a manager for every condition, each treating a separate person. There was no one to coordinate the care, to look at the whole person and really engage that person in treatment.”

Schooled by these earlier failures, major health insurers are investing in a new generation of “population health management” (PHM) technology. PHM programs continue to depend on the latest predictive analytics, but now leverage a range of knowledge, tactics, and technologies to engage individuals, track and record progress, and generate more positive and better-measured outcomes.

Health plans look for consumer engagement

“Medicine is catching onto the idea that behavioral health approaches can engage the individual and promote the positive behavior changes needed to make medical interventions successful,” says Farook. He notes that two-thirds of healthcare spending, which now averages about $9,000 per person in the U.S., goes to treat individuals with chronic diseases. Most of the depression that occurs in this group is treated by primary care physicians with medication-a good start-but, Farook adds, “these people aren't getting exposed to behavioral health services, which provide a better long-term outcome.”

The key to the most effective PHM approaches, he says, is in identifying chronically ill individuals and linking them up-as part of their chronic disease treatment-with behavioral health resources. For any plan or large employer, the analytics process is the same: Run mountains of data to segment the population into groups-well, at-risk, or chronic, for example. For chronic groups, plans typically adopt more engaging, interactive approaches, such as the use of personal coaches-trained in behavioral health-to establish relationships, then engage the individual regularly through calls or more current technology-such as e-mails, text messages, and mobile applications-to encourage participation, monitor progress, and “keep them in the loop.”

“I'm seeing a lot of this ‘med-psych integration,’” says Farook. “There's an understanding that health coaching models can make the difference in cost containment.” As managed behavioral health organizations develop new products and services aimed at both chronic and well populations, Farook says that businesses like his are moving beyond analytics and measurement to expand the “back-office” capabilities needed by payers and plans to support engagement through the latest communications devices.

Expand data collection, employ decision supports to improve outcomes

As payers and plans refine their analytical tools and customer engagement tactics, behavioral health experts, like those at Polaris Health Directions (Langhorne, Pa.) offer tools that enable behavioral healthcare providers to individualize treatment, improve outcomes, and increase program effectiveness.

Linda Toche-Manley, PhD, vice president at Polaris, says that providers can improve outcomes by:

• Collecting additional client diagnostic data;

• Providing that data to clinicians so they can select and plan the optimal service mix for treatment; and

• Making “risk-adjusted” evaluations of the quality and effectiveness of programs that target specific populations.

She cites a typical example: An individual reporting severe depression presents and requests treatment. He is asked to fill out personal history forms. But he does so selectively, omitting information about a chemical dependency problem. “This is not unusual,” says Toche-Manley, who adds that “the initial course of treatment is typically determined by the information [caseworkers] are given, in combination with the caseworker's educational backgrounds and experiences in the field.” As a result, “rarely are recommendations based on empirical data that is linked to potential outcomes.”

She suggests a different process, which supplements the forms with a computer-based personal assessment-a tool more likely to get a truthful result. With the individual's permission, his wife and family members also complete an assessment. This additional information on the individual's behavior, risks, and strengths-backed by the use of a clinical decision support tool-provide the caseworker with greater insight, measures of the relative severity of the individual's symptoms, and a set of evidence-based treatment approaches matched to the individual's symptoms. Every few weeks during the course of treatment, the individual takes computer-based assessments that track his progress.

So, what's the difference? Toche-Manley points to three important differences. First, she says “the voice of the patient, as well as standardized data, are used to build the treatment plan.” Second, additional information regarding the individual's risks and strengths-information that empirically affects long-term prospects for recovery-is documented. And third, the cycle of measurement and feedback drive better communication and progress evaluation between the individual and the clinician.

In addition to impacting the clinical process, Toche-Manley says that significant program-level improvements can be realized by analyzing key individual risks, strengths, and other factors. “Many program directors wonder, ‘What programs work best with what types of clients?’ Treatment populations are not homogeneous, and what succeeds for one client may fail for another,” she explains, citing a multi-county study of the effectiveness of wraparound care program outcomes for nine types of youth. Through analysis of various combinations of clinical and service data, she says that “what does and does not work for specific youth and adults began to be revealed.” When applied program wide, these analytical insights are improving outcomes by over 30 percent.

“Real cost savings are often realized by better initial matching of services to specific clinical subgroups, and not by elimination of service types,” says Toche-Manley. “Systems can afford to give the best services-even the most costly ones-when their service decisions are supported by empirical observations that probable success is high.”

Behavioral Healthcare 2010 July-August;30(7):22-23