Reprinted with permission from Managed Care and Cancer
Volume 11, Number 5/ May 2002

The Florida Blues' Experience in Improving Cancer Management
By David L. Teitelman and Frederick C. Lee

Cancer patients account for as much as 15% of all medical expenses in a health plan. Payers, like Blue Cross Blue Shield of Florida (BCBSF), spend 6% to 9% of all medical expenses on cancer treatment. Ironically, treatment features common to cancer are compatible with the principles of managed care, yet they rarely are implemented by health plans.

In 1996, BCBSF and Quality Oncology, Inc. (QO), a cancer disease management firm founded in South Florida in 1993, began exploring a pilot program to address cancer management. In 1999, BCBSF engaged QO to implement the pilot program for 225,000 commercial and 55,000 Medicare Risk lives in BCBSF’s managed care program in Dade, Broward, and Palm Beach counties. It’s success, described below, has led to a statewide rollout of the program.

PROGRAM GOALS AND OBJECTIVES

The goals of the cancer program for BCBSF included: improve clinical outcomes and quality of care; increase member and provider satisfaction; support accreditation efforts (eg, NCQA, HEDIS, QISMC); and manage medical costs.

Additionally, the BCBSF/QO joint management team initiative sought to achieve the following objectives: (1) Reduce unnecessary inpatient utilization when an appropriate, lower-cost setting is available, (2) reduce the readmission rate, (3) reduce the number of inappropriate laboratory and diagnostic tests, (4) increase hospice utilization, (5) decrease complications associated with disease progression and treatment, (6) reduce variation of treatment patterns through the use of clinical practice guidelines by physicians, (7) ensure appropriateness of care, and (8) identify when spending money will save money.

PROGRAM APPROACH

Quality Oncology and BCBSF designed the following approach to meet the goals and objectives:

• Precertify all cancer services using nationally accepted, evidence-based guidelines.

• Monitor and maintain regular collegial interaction with local oncologists.

• Expand utilization to a full continuum of care.

• Improve management of cancer infusion therapies, including chemotherapy and supportive care drugs.

• Better monitor and influence care provided in tertiary care centers.

• Increase appropriateness of end-of-life treatment.

• Introduce measurement of cancer costs and processes.

• Deploy cost and quality analyses to allow physician and hospital profiling.

• Report on cost and quality by comparing with a historical baseline.

• Plan changes based on documented evidence (or lack thereof) of impact.

PROGRAM IMPLEMENTATION

Quality Oncology hired a client-specific team of five oncology-certified nurses and five licensed practical nurses.  A dedicated toll-free telephone line was established for the Florida Blues, and a South Florida provider-relations orientation to QO was initiated in the winter of 1998.  BCBSF dedicated three staff and portions of other managers’ time to the program.  Interactions between the two firms were significant from the program’s inception.

The implementation team established a baseline of historical cancer experience, using the 24 months prior to the start date (February 1, 1997 to January 31, 1999).  The baseline was created from claims data, whereby cancer patients were

'The team established meaningful quality improvement metrics by which Quality Oncology would be measured, including patient satisfaction.'

identified using ICD-9 codes, and then their cancer coded claims and related cancer expenditures were aggregated.
The team also established meaningful quality improvement metrics, by which QO would be measured, including a focus on patient satisfaction, provider satisfaction, provider adherence to evidence-based cancer treatment guidelines, and reductions in unnecessary readmissions and emergency room visits.

ENROLLMENT

BCBSF had approximately 8,000 members with a definable cancer condition in each program year. Quality Oncology case-managed 446 cancer patients in program year 1 and 775 in program year 2.

The tumor type specificity of the population showed almost no variation from what QO has seen in other client data sets, with prostate being the most common Medicare cancer and breast being the most common cancer affecting those under age 65.

The five most common cancers (breast, prostate, lung, colorectal, and lymphoma) absorb 65% to 70% of the total cancer expenditures and affect a comparable percentage of the cancer patients.

The opportunity for targeted, tumor-specific management is deduced from a heightened disequilibrium between percentage of patients and percentage of dollars spent in certain cancers. For instance, lung cancer is a tumor type for which the dollars expended are disproportionately greater than the number of patients served. Colorectal cancer, lymphoma, digestive cancer, and leukemia are other tumor types in which this disequilibrium occurs.

CHANGE IN INPATIENT USE

The most consistent impact of the program has been to divert or eliminate unwarranted hospital admissions. A decline in admissions can be attributed to a variety of initiatives:

Inpatient chemotherapy sessions are moved to the outpatient setting when appropriate.

Patients undergoing chemotherapy are closely monitored for nausea and vomiting, preventing admissions for dehydration.

Patients are queried routinely about their pain status, and their supplies of pain medication are scrutinized to prevent them from exhausting their medication caches, leading to unnecessary emergency room visits.

Planning around terminal issues is addressed early on in the treatment process, so that unwanted hospitalizations at the end of life are pre-empted (see Figure 2).

Based on QO’s efforts with BCBSF cancer patients and a sentinel effect that influenced South Florida oncologists, the number of hospital days per case declined 11.1% for the entire commercial cancer population and 14.6% for the entire Medicare Risk cancer population from baseline to program year 2 (Figure 3). As is apparent for the figure, the successful impact on hospital days was sustained in program year 2 for both populations.

BCBSF believed that a series of performance standards that directly influenced quality of care had to be integrated into the program’s design. Both parties were intent on defining measurable standards that would improve quality for the members. The measures were slightly refined between program year 1 and program year 2.

Both parties were strong proponents of evidence-based guidelines as a way to improve quality by reducing variation. A standard of 90% of patients being treated within guidelines was set, and the standard was exceeded in both years: 91% in year 1 and 91.4% in year 2. As a sign of the effort required to prompt 
physicians into adhering to guidelines, medical director interventions increase 42% in year 2.

Patient education efforts were initiated. Quality Oncology believes that getting cancer patients to become more knowledgably about their conditions and associated side effects yields better results for both patients and payers. Empowered patients can use hospitals more prudently.

The program sought to keep readmissions within 30 days of discharge below 16% of all discharged patients. In year 1, this standard was met, with 14.8% of discharged patients being readmitted in fewer than 30 days; year 2 results slipped to 16.4%.

Emergency room use was targeted at fewer than .086 visits per cancer patient. The program missed the target by .002 in year 1, but significantly beat the same mark in year 2 with .043 visits per cancer patient

An important measure of the program’s success, in both BCBSF and QO’s eyes, was the managed cancer patients’ responses to a survey. In 1998, QO developed and validated an 11-question, cancer-specific patient satisfaction survey. The questionnaire is sent out 30 days after “graduation” from the case management program.

Most graduates of the program enter a period of remission, a point where QO’s weekly involvement is no longer cost-effective to its clients. Unfortunately, some “graduates” of the program are deaths, and in those cases, their loved ones are surveyed.

On the 1 to 5 scale of satisfaction across all 11 questions (with 5 representing the highest satisfaction), QO achieved an average of 4.13 for the 45% of people who received and returned their surveys.  QO’s highest average score was associated with providing meaningful information on a patient’s cancer; its lowest average score reflected its efforts to better coordinate patient care.

For year 2, the score rose to 4.38, but the standard was slightly modified.  We set as a goal that 90% of those involved in the cancer care program would rate their experience as good (4) or excellent (5).  Actual results fell shy of the goal, with 85% of the respondents judging their experience as good or excellent.

Physicians were asked to rank their satisfaction with their experience working with the cancer care program, on the whole, on a scale of 1 to 5, with 5 being the most satisfied.  The goal for the average score for physician satisfaction was set at greater than 4.1.

In year 1, the average score of the responding physicians was 3.51.  In year 2, it rose to 4.11, demonstrating that, over time, physicians tend to become less resistant to the service and more accepting of its value in assisting patients and supporting doctors (see Figure 4).

In year 2, a second assessment of provider satisfaction was implemented, based on use of a physician-only telephone helpline.  The goal was 95% of calls answered in 20 seconds.  The actual result was only 70%, with an average answer speed of 33 seconds. 

Over the past year, many health plans have observed increases in their chemotherapy costs/service in the 25% to 35% range.  Use of injectable drugs for treating neutropenic fever, fatigue, and pain has also increased rapidly.

BCBSF had addressed the use of injectable agents prior to its engagement with QO, by instituting a drug replacement program in which patient-specific chemotherapy is drop-shipped to the physician’s office.

 QO was then given the task of precertifying all authorizations for chemotherapy and supportive care injectables to determine the appropriateness of treatment regimens and course of treatment.  Although activity continued to soar at a double-digit pace, BCBSF did observe a decline in precertification requests, and billing procedures.

On a services per thousand basis, chemotherapy infusion and supportive care injectable requests dropped 10% vs. the baseline for the commercial population, but increased 5% for the Medicare Risk group.  Leuprolide (Lupron), a hormone injection for men with prostate cancer, was the number one injectable drug, accounting for 20% of the entire injectable budget.

Based on the considerable success of the program not only in minimizing hospital use and costs but also in eliciting significant satisfaction form the membership, BCBSF decided to expand its commitment to the program statewide.

CONCLUSION 

Both the client and the vendor characterized the pilot program as a success.  Quality Oncology further refined its service with more frequent provider relations activities, more attention to tertiary centers, and a new product introduction for managing the cost of injectable drugs.  Both parties were pleased with the quantification of results, a process that is foreign to most health plans.

Climbing satisfaction scores from the surveyed physicians in the South Florida market suggests that this program succeeded in overcoming predictable initial resistance from doctors, which was more a function of not knowing QO’s role than of objecting to the value-added services that were being provided to cancer patients.

Quality Oncology prevailed in proving to local physicians that its services were supplemental and appreciated by cancer patients and not further bureaucratic measures impeding the flow of patients and their information.