Outthinking the odds: insurer has predictive computer algorithm up its sleeve

Preventing unnecessary admissions is now a task being taken up by more than the government and hospitals. One insurer—Philadelphia-based Independence Blue Cross—is joining the preventable readmissions fight with a unique approach to the problem. The company is using data and a complex algorithm to identify high-risk patients so that they can be selected for heightened service in an attempt to prevent costly readmissions.

Finding risks

According to a story from a Kaiser Health News (KHN), Somesh Nigam, the Chief Informatics Officer for Independence Blue Cross,  said the company is trying to identify patients who are “likely to be hospitalized in the next three months.” The idea is to find the patients who are just on the edge of requiring hospitalization.

Algorithm

To accomplish its goal, Independence Blue Cross runs a complex algorithm on a mountain of data. The data includes billing claims, lab readings, medications, height, weight, and family history, as well as information about a patient’s neighborhood and poverty rates. The quantity of information is significant. Nigam said that the health care data is comparable to all of the data found in five Wikipedias. The idea behind the algorithm is basic. It analyzes the data entered and then generates a score for a patient based upon the patient’s estimated risk of hospital readmission.

Coach

The insurer then assigns a “health coach” to patients with high scores. The coach works with the patient to identify which services might assist that high-risk patient, in order to prevent even more costly services. One patient, John Lovine, experienced the benefit first hand when a health coach assisted him in setting up an appointment with a visiting nurse as an alternative to a costly hospital visit for bloodwork. According to Nigam, the program is working and the insurer is already seeing dropping hospitalization rates in its coverage region. Independence Blue Cross has identified 18,000 patients for the health coach program and has seen a 40 to 50 percent drop in expected hospital admission rates for patients with congestive heart failure.

Concerns

Despite the successes, a Harvard Bioethics Law Professor, Glen Cohen, expressed some concerns to KHN about the ethical implications of the algorithm. Cohen said the system raises questions about whether patients should have the right to opt out or should have to affirmatively opt in. Although Independence Blue Cross says its follows federal privacy rules, Cohen says the appropriate handling of the information is still something that is up in the air. Regardless of its final form or long-term implications, the algorithm represents a novel attempt to get a step ahead of health care costs, a step that is likely to benefit patient health as well as the insurance company’s bottom line.