Best practices reduce costs of chronic care; northeast MCO uses predictive modeling, simple data mining and enhanced provider relations to manage the medical costs and personal impact of diabetes

Best practices reduce costs of chronic care; northeast MCO uses predictive modeling, simple data mining and enhanced provider relations to manage the medical costs and personal impact of diabetes

The costs of chronic care have placed a strain on the current healthcare system, accounting for approximately 80 percent of total healthcare costs.

In addition, the upswing in diabetes in America is a troublesome one. According to the Centers for Disease Control and Prevention, diabetes affects more than 18 million Americans and is now the sixth leading cause of death in this country, with Type 2 diabetes accounting for up to 95 percent of all cases.

The treatment regimen for diabetes is complex and challenging to the average patient. It relies on expensive drug treatments, and, if the treatment regimen is not followed closely, even more dangerous and costly complications such as heart disease and hypertension may develop. Right now, federal estimates are that diabetes costs the U.S. $132 billion a year, $92 million of which are in direct medical costs. Just two years ago, the healthcare cost of treating a diabetic was about six times that of treating a person without the disease, and those numbers can only increase.

One of the most difficult barriers to overcome when trying to engage diabetic members in taking control of their health is influencing behavior change. Often individuals with chronic conditions such as diabetes do not understand how to manage their conditions and lack the information necessary to be compliant with treatment plans.

Applying Best Practices

Oxford Health Plans, headquartered in Trumbull, Conn., offers health plan products to employers and individuals in Connecticut, New York and New Jersey, and serves more than 1.55 million members via networks that exceed 57,000 physicians and 227 hospitals in the Tri-state area. In response to the growing number of individuals with chronic conditions, Oxford implemented a pro-active program to help curb these costs by helping members with chronic conditions receive the most appropriate care.

The program, called the Oxford Best Practices Program (BPP), rolled out in 2003 with the Diabetes Best Practices Network (DBPN). The DBPN program was created to address the health needs of plan members. Through it, Oxford works with the member’s primary care physician (PCP) to help him or her deliver the best possible care and resources to the member and, thus, to effect behavioral change in diabetic patients. The program is designed to reach those individuals who are not engaging in any of the appropriate activities, such as healthy diet, exercise and regular visits to specialists, to control their diabetes.

Qualifying Physicians

Oxford identified 28 high-quality endocrinologists to participate in a pilot program around the treatment of members with diabetes. The endocrinologists working in this program have either achieved, or will apply for, recognition from the National Committee for Quality Assurance’s (NQCA) Diabetes Physician Recognition Program, a voluntary program cosponsored by the American Diabetes Association that assesses a physician’s practice in attaining key measures of quality diabetic care. Additional criteria used in creating this network included: 1) the physician’s geographic location relative to areas with high concentrations of diabetic members; 2) the physician’s ability to accept and evaluate new patients in a timely manner; and 3) the physician’s expertise in treating diabetes.

Following initial identification, Oxford’s profiling software vendor, Integrated Healthcare Information Services (IHCIS) of Waltham, Mass., and Oxford’s Health Care Economics department further screened the physicians for eligibility and to evaluate each physician’s financial and clinical efficiency. Physicians who met all DBPN standards were invited to participate in the diabetes pilot program. To date, 28 endocrinologists in the Bronx, Kings, Nassau, New York, Queens, Suffolk and Westchester counties are participating in the pilot study.

After creating the DBPN, Oxford began identifying members who had not visited an endocrinologist in more than a year and who also exhibited an HbA1c of greater than 9.5. Once these members were identified, Oxford began reaching out to each member’s PCP with information about the DBPN. Physicians interested in enrolling their patients in this program can refer their patients to a DBPN endocrinologist or, if the physician chooses, can have Oxford contact the patient directly and make referrals.

For the initial phase, Oxford identified more than 250 diabetic members in six New York counties–New York, Queens, Richmond, the Bronx, Nassau and Westchester counties–who had HbA1c levels significantly above the standard for diabetic control. These members also had not seen endocrinologists in a significant period of time.

In addition to the BPP for diabetes, Oxford also created the Living with Diabetes program, which includes multiple interventions for members and providers, including educational materials, seminars and disease-specific exam reminders. Currently, there are 50,000 members enrolled in this program

Predictive Modeling at Work

Oxford uses Impact Pro, predictive modeling software designed by IHCIS. The software analyzes claims data against lab results data to identify arenas where the member is not following his physician’s prescribed treatment or is otherwise at risk. The software helped to enable a strategic shift in how Oxford treats chronic care patients because it optimizes the use of claims data to predict those members at highest risk, rather than looking at already-occurred adverse outcomes and then determining a remedial plan.

The predictive tool is a multi-dimensional, member-based reporting tool that analyzes clinical information across the range of providers treating the patient and establishes a score that represents the risk of an impending hospitalization or other negative event. Oxford identifies the members who are most at risk of complications and initiates the process of contacting them for permission to refer them to DBPN providers.

Using a rules-based model, Impact Pro creates markers of risk on a monthly basis that can be predictive and also may lead to a greater clinical understanding of why a diabetic patient is at high risk. The software is based on the “episodes of care” approach, which uses data about a patient’s clinical episodes including prior utilization of services and prescription drug use as risk markers.

PCPs with high-risk diabetic members in their panels received lists of endocrinologists in New York who received recognition from NCQA for the quality of care they provide to diabetics. Oxford asked the PCPs to encourage their diabetic members to make appointments with NCQA-recognized endocrinologists to discuss diabetes management. The program is entirely voluntary.

Proof of Effectiveness

Over the course of the BPP, Oxford has mailed more than 1,800 letters to PCPs and 2,300 letters to members regarding the best practices network. In addition, care managers have made more than 7,000 calls during a one-year period to follow-up with members.

To date, more than 380 high-risk diabetic members have been referred to DBPN endocrinologists. Of those members with confirmed visits to specialists, an average 8 percent decrease in their HbA1c levels has occurred. According to national studies, for every 1 percent reduction in HbA1c, the relative risk for microvascular complications decreased by 37 percent, diabetes-related deaths decreased by 21 percent, and heart attacks decreased by 14 percent (heart attack reduction was of borderline statistical significance). In addition, Oxford has received reports from DBPN endocrinologists that more than 117 initial member visits, 24 second visits, five third visits and two fourth visits have been recorded. According to data received from these specialists, more than 60 percent of participating members have improved their HbA1c results by their second visit.

Oxford is now working to expand this program to include other chronic conditions that can benefit from following a best-practices mode of intervention.


Alan M. Muney, M.D., M.H.A. Executive Vice President and Chief Medical Officer Oxford Health Plans Trumbull, Conn.


Impact Pro Integrated Healthcare Information Services Waltham, Mass.

For more information about Impact Pro from Integrated Healthcare Information Services,

COPYRIGHT 2004 Nelson Publishing

COPYRIGHT 2004 Gale Group