FLU SHOTS FOR OLDER ADULTS

Collaborating to Reduce Missed Opportunities

Every year in the United States, an average of 110,000 hospitalizations and 20,000 deaths are related to influenza; 90 percent of these deaths occur in the elderly. In the elderly population, the flu vaccine can reduce the risk of hospitalization by up to 60 percent, and the risk of death by up to 80 percent. [1]

In 1995, the state's Medicare Peer Review Organization (PRO) identified improvement in the flu vaccine rate among the Medicare population as an important initiative. It invited managed care plans statewide to join in a collaborative effort to address the issue.

With a rapidly growing Medicare population, and a high incidence of both inpatient and out-patient claims for chronic cardiopulmonary diseases, the plan felt that increasing influenza immunization rates could have a significant impact on the burden of illness for its membership.

As early as 1994, the plan's marketing department had sent postcards reminding members of the need for flu shots, but it was not clear if the right population was being targeted, and there were no measures of the postcards' effectiveness.

The plan decided to join the statewide project spearheaded by the PRO to improve flu shot administration among its seniors.

The plan targeted all members of its Medicare risk product who had no more than one break in enrollment of up to 45 days. (The continuous enrollment criterion was selected to ensure that members would receive all information regarding influenza vaccine, and would be eligible during the time period when vaccine was administered.)

The performance measure selected was the percentage of the targeted population who had received a flu immunization between September 1 and December 31 of the reporting year. The plan used claims and encounter data, looking for either a CPT-4 code of 90724 or an ICD-9-CM code of V04.8.

ADMINISTRATIVE DATA One of the reasons the plan used administrative data was to conserve resources. Staff was not available to abstract medical records for this project. Careful planning up front allowed the plan to collect data on both missed opportunities and performance rates with a single data pull.

Baseline performance, based on 1995 data, showed an immunization rate of 43 percent. A plan involved in the statewide collaborative project had achieved a rate of 80 percent. This plan used that rate as its benchmark. The plan adopted a goal of 60 percent from the U.S. Public Health Services' Healthy People 2000.

A "flu team" composed of the plan medical director, provider representatives, pharmacy, claims, IT and QI staff advised the Quality Improvement Committee throughout this project. Possible barriers were identified during brainstorming sessions with the PRO and representatives from health plans statewide. State rates and experiences from all the health plans drove the analysis. The flu team felt that a lack of physician involvement was a prime barrier. Although some steps, such as the distribution of an adult immunization practice guideline, and reimbursing flu shots outside the capitation payment, had already been taken, the plan felt additional measures could be implemented.

The plan used its transactional data to identify members with one or more PCP visits, but no flu shot. Analysis showed that 33.3 percent of the targeted population had one or more missed opportunities.

The plan implemented member education with articles in the member newsletter, reminder postcards, and personal letters as a follow-up to the postcards. Information packages were mailed to the PCPs, and included office posters, buttons and chart stickers.

A SINGLE MESSAGE The involvement of the state's PRO helped facilitate clinician buy-in. The PRO communicated directly to the physicians with its information package. As a result, one unified mailing of materials with a single message went to each practitioner. This avoided confusing or conflicting materials from multiple health plans.

The first remeasurement, based on 1996 data, documented an immunization rate of 52.3 percent, a statistically significant improvement using the z-test (p value <0.05). However, the measure failed to meet the plan's performance goal or match the state's benchmark performance.

An analysis of the plan's missed opportunity rate showed a decrease to 27.7 percent, also statistically significant. The plan felt this showed that PCPs were becoming more proactive in immunizing seniors.

Flu vaccine rates by individual offices were placed on a scatter chart. The chart contained lines indicating the unweighted mean, and rates that were one and two standard deviations below the mean. Three offices showed performance two standard deviations below the mean.

ANALYZING DATA WITH CAUTION The three outlier offices had very small numbers of Medicare enrollees. With transactional data, it was impossible to determine which patients had shots elsewhere, refused the shots, or had contraindications. Caution needed to be exercised before jumping to conclusions, such as labeling the practice sites as poor performers. The plan decided to track these three offices in subsequent years to see if they remained outliers. None was an outlier the following year.

Physician newsletters continued to inform clinicians of the importance of flu shots. The state PRO once again coordinated mailings of a flu vaccine package containing posters, buttons and other reminder items. Office managers were sent mailings (with copies to the PCPs) showing the scatter chart with results of the 1996 campaign, and lists of specific patients who were Medicare members, and therefore, candidates for flu shots.

The second remeasurement, based on 1997 data, showed an immunization rate of 59.3 percent.

This increase was highly statistically significant (p value = 0.001). However, the plan had not quite met its goal.

Missed opportunity analysis showed a decrease to 13.3 percent. This was a statistically significant decrease, and was felt to reflect the positive effects of the feedback and education of providers.

The office rates were once again placed on a scatter chart. It was noted that 22 office sites, all with small numbers of Medicare patients, had submitted no claims for influenza vaccines. Since none of these offices was a repeat outlier, they were, once again, tracked to see if they would continue to be outliers.

The plan mailed a brief member survey (included in QP Tool ) to seniors who had no transactional data demonstrating a flu shot. This survey asked if, indeed, the member had received a flu shot, and if so, where. If the members had not received shots, they were asked the reason why. The overall response rate for members surveyed was 56.3 percent.

Survey results revealed that 46.3 percent of responders had not received the vaccine. The top two reasons? An equal number (42 percent) reported previous reactions, and fear of side effects. In addition, 4.5 percent stated they forgot, and 1.3 percent were unaware of the need for the shot.

The survey also showed that 53.6 percent of respondents had received the shots, either at their PCPs' offices or at community sites.

Based on the survey and calls placed to the member services department, it appeared that many members had their flu shots at community sites sponsored. by one of the community's home health agencies. Although these sites offered ease of access to the seniors, they charged $10 for the vaccine. This represented a significant financial burden for some members. The plan decided to collaborate with the home health agencies and reimburse them for the provision of vaccines at community sites. The plan required the agency to submit documentation of the immunization on an immunization record (included in QP Tool ). This record not only helped the plan identify immunized members) but also was used to communicate with the PCPs.

Encounter data issues were addressed for members who stated they had flu shots at their PCPs' offices, but had no record in the claims system.

The PRO continued mailing influenza kits to PCPs, and announcements of the package mailings to office managers.

The plan mailed the PCPs their 1997 immunization rates, the scatter chart showing their performance and lists of members needing immunization. Providers were also mailed a list of community sites where their members could receive flu shots.

The member newsletter contained an article addressing myths and facts related to the vaccine.

All Medicare members received reminder postcards.

1998 and 1999 data showed rates of 55.8 percent and 56.6 percent. While these represent slight decreases, the plan has been able to sustain its improvement in immunization rates.

[1] - "Life Savers for Seniors," In CMRI website [page on website of CRMI]. San Francisco, CA 2000 [cited December 13, 2000]. Available from www.crmi-ca.org; INTERNET.