FLU SHOTS FOR OLDER ADULTS

Measuring the Effect of Specific Interventions

Health plans embarking on quality improvement activities often implement numerous interventions, and are then at a loss to determine which were most effective. Not so with this plan. It stratified its targeted population into discrete groups, each with a different set of interventions. By rigorously analyzing the results from each group, it was able to determine which interventions were most effective in encouraging its elderly members to obtain flu shots.

Influenza viruses have been responsible for 18 epidemics between 1957 and 1985, and more than 40,000 excess deaths occurred during each of the two most recent epidemics. [1]

Eighty to 90 percent of these deaths were attributable to influenza and pneumonia among persons 65 and older. [2]

During the period from 1993 to 1997, this plan's Medicare population grew from 1,783 to 142,416.

Based on the knowledge of the importance of influenza immunization in its rapidly growing membership, the plan embarked on a study to evaluate the rate of compliance with influenza immunizations among its over-65 population.

The plan initially targeted all members of its Medicare risk plan. In 1995, it expanded the targeted population to include commercial members over age 65.

The performance measure selected was the percentage of the targeted population who had received flu immunizations between September 1 and January 31 of the reporting year. The plan used claims and encounter data, looking for a CPT-4 code of 90724. The plan calculated separate rates for its Medicare and commercial populations.

Baseline Medicare performance, based on September 1993 to January 1994 data, showed an immunization rate of only 11 percent. This was substantially less than the Healthy People 2000 goal of 60 percent, which the plan used as its performance goal.

Based on the Behavioral Risk Factor Surveillance System (BRFSS), a national telephone survey developed by the CDC, benchmarks were selected. In 1995, the plan chose the highest performing state with a rate of 70 percent. [3] In 1997, the highest performing rate was 74.4 percent. [4]

During the 1995 - 1996 flu season, a baseline measurement 0f 19.1 percent was reported for the plan's commercial members over age 65.

An influenza committee was established early on in the activity. The committee was composed of the quality manager, Medicare manager, a representative from medical management, a marketing representative, a biostatistician, and a medical director. This group identified two initial barriers:

• Lack of member awareness: the plan had never undertaken any educational effort targeting flu shots in its relatively new Medi care product. With influenza immunization left to the discretion of individual physicians, it was felt that many members were unaware of the need for annual flu shots
Economic barriers: although health plan policy was that flu shots were free (no co-payment), members often did not know this. Similarly, even though the plan reimbursed flu shots on a fee-for-service basis above the capitation fee, many providers were unaware of this policy

Both of these barriers were addressed in the 1994-1995 educational initiative.

The plan implemented a program that encouraged members to receive preventive care. Preventive care reminders were sent for a number of services. (The program is summarized in QP Tool .) All members age 65 and older received a mailing reminding them of the need for a flu shot. (Sample mailings are included in QP Tool and QP Tool .)

The plan medical director sent a letter to approximately 7,000 PCPs. It informed them of the advantages of influenza immunization and of the member education program. It included a fact sheet detailing billing and coding of flu shots, and outlining a process for implementing an influenza immunization program in the office. (A copy of the mailing is included in QP Tool .)

The plan supplemented these efforts with articles in member and provider newsletters.

The first remeasurement for the Medicare population, based on September 1994 - January 1995 data, documented an immunization rate of 20.8 percent. This was a statistically significant improvement, using a chi square test (p<.01). However the plan failed to meet its performance goal.

The plan reviewed the literature to identify possible interventions for, and barriers to, influenza immunization. Positive predictive factors derived from the literature [5] included:

• Past history of vaccination
• Physician or nurse recommendation
• Expressed intention to follow vaccination recommendation from physician/nurse

Negative predictive factors derived from the literature included:

• Not believing the vaccine prevents influenza [6]
• Not vaccinated in previous year
• Never been vaccinated
• Felt sick after past influenza vaccination [7]
• Fear of side effects of vaccination [8]

To address these factors in its own population, the plan developed and conducted a telephone survey. It administered this survey to a random sample of newly enrolled Medicare members receiving a welcome call from the plan.

USING THE LITERATURE One of the strengths of this activity was the degree of research that went into understanding the clinical and psychosocial issues of flu vaccination. The plan discovered both factors that tend to encourage immunization, as well as those that tend to discourage it. The plan didn't stop there. It used the data from its literature review to design member surveys to look at the importance of the identified factors in its own population.

This survey revealed that 46.1 percent reported receiving a flu shot. Of this group:

• 74.7 percent received them in their doctor's office
• 19.5 percent received them in a clinic
• 5.8 percent were immunized in another location

WELCOME CALLS TO SENIORS A variety of barriers may be related to poor understanding of a plan's benefits. This may be particularly true of a Medicare population. This plan had an existing process for extending welcome calls to its newly enrolled Medicare members. These outbound calls reached 80 percent to 90 percent of newly enrolled members. The benefits and protocols of the health plan were explained to seniors.

Of the 53.9 percent who reported not receiving a flu shot:

• 55.1 percent did not think it was important
• 34.6 percent were afraid of side effects or getting sick
• 2.7 percent said their doctor did not tell them to get the shot
• 2.7 percent stated other reasons

Based on the literature search and member survey, the plan selected three additional barriers to be addressed by the activity:

• Lack of perceived importance of vaccination
• Side effects
• Lack of physician recommendation

The influenza committee designed an intervention that would allow the evaluation of different strategies in encouraging immunization among its Medicare members. It randomly assigned seniors to one of four groups:

• A control group that received no mailing
• A group receiving the regular mailing of a letter and postcard reminder
• A group receiving a mailing of a letter, a postcard reminder, and a second letter with member testimonials (included in QP Tool )
• A group receiving a mailing of a letter, a postcard reminder, and an incentive (an entry into a drawing to win one of two $500 gift certificates)

The plan continued its program of provider mailings. These mailings included statistics on influenza, a copy of member letters, and billing and coding instructions for influenza immunizations.

EVALUATING SPECIFIC INTERVENTIONS The plan thoughtfully evaluated a series of different interventions, and compared them to a control group with no interventions. This not only helped the plan to identify and expand the most effective interventions, it allowed the plan to discontinue less effective ones.

One medical group began to place phone calls to its senior members, encouraging them to make appointments for flu shots.

The second remeasurement, based on September 1995 - January 1996 data, showed an immunization rate of 32 percent among Medicare members. This increase was statistically significant. This was the first year that the commercial population over 65 was included in member education mailings; the baseline rate of influenza vaccination for this group was 19.1 percent.

The plan evaluated the immunization rates for the four groups of Medicare beneficiaries with different interventions (Table 1).

The control group was found to have a statistically lower chance of immunization than the other groups. The groups with the regular mailings, and those receiving the testimonials were statistically higher than the grand mean, but were not statistically different from each other. The group receiving the incentive was found not to differ statistically from the mean.

Based on this analysis, the plan felt that the regular mailings and the mailings with testimonials were the most effective interventions.

In addition to previously identified barriers, the plan noted the existence of cultural barriers (4 percent of the Medicare population had Chinese as their primary language). It also noted that members who were hospitalized were often not being vaccinated. It continued to be concerned about the need for a strong intervention regarding physician recommendation of the flu shot.

CULTURAL SENSITIVITY The success of this plan's interventions owed a lot to the cultural sensitivity of the influenza committee. The most obvious example was its recognition of a potential language barrier, and its response with the development of outreach letters in Chinese. Less apparent, but just as important, was its willingness to explore the "culture" of senior citizens through a literature search and a telephone survey, which uncovered widespread disbelief in the effectiveness of flu shots and fear of side effects.

The plan continued its member mailings, with postcard reminders and testimonials. It sent out over 3,600 letters in Chinese.

One medical group continued its phone call reminder system. It encouraged members to make appointments, and provided after-hours appointments where needed.

Two other new interventions were piloted. 3,000 members received personalized letters with their physicians' names and addresses. Additionally, patients admitted to one of 10 study hospitals were the subjects of an intervention designed to increase immunizations among hospitalized patients. The medical director faxed immunization recommendations and lists of patients to all PCPs having patients (identified from the prior authorization system) hospitalized in one of these 10 hospitals.

The plan continued its provider mailings and published articles in provider and member newsletters. (An example is included in QP Tool .)

The third remeasurement, based on September 1996 - January 1997 data, showed an immunization rate of 36 percent among Medicare members, and a rate of 24.5 percent among commercial members over age 65. Both of these rates represented statistically significant increases over the previous year.

Evaluation of the Chinese language letters and the personalized letters with the PCP names and addresses showed that these two interventions resulted in immunization rates significantly greater than the mean.

Measurement of immunization rates of the members receiving phone calls from their medical groups showed the highest rate of immunization - 51 percent.

The results of the pilot study of vaccinating hospitalized members were disappointing -only 2 percent of these members received flu shots. When physicians were surveyed about why shots were not given, common answers included early discharges with short lengths of stay, and the attitude that the acute care setting was not an appropriate one for immunizations.

The plan identified the continuing importance of three barriers:

• Lack of physician recommendation: the plan expanded the numbers of members receiving customized letters with their PCPs' names and addresses. It sent member lists to PCPs in six medical groups
• Lack of vaccination of hospitalized members: the plan concluded that it did not have enough influence over its PCPs to mandate immunizations during inpatient stays
• Immunization by nonparticipating providers: the plan adopted a policy of paying for influenza immunizations on claims from any provider

The fourth remeasurement, based on September 1997 to January 1998 data, showed immunization rates of 43 percent among Medicare members, and 52.2 percent among commercial members over age 65. Although the plan had not met its goal, both measures showed statistically significant improvements from the previous year; the plan established a significant linear trend in improving influenza vaccination rates in both target populations.

The plan continued its previous interventions, continuing to address barriers in physician recommendations with mailings to members and PCPs. It added another high-risk group to its targeted population: diabetics under 65 years of age. It initiated a targeted mailing to these members.

This activity allowed staff and departments within the plan to participate in an activity that was positive for members. Staff reported that they felt like the "good guys," noting the positive effects on morale from engaging in this effort to keep their members well.

[1] - Department of Health and Human Services, Centers for Disease Control and Prevention, "Recommendations of the Immunization Practices Advisory Committee (ACIP) Prevention and Control of Influenza," Morbidity and Mortality Weekly Report, 39 no. 20 (Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, May 23, 1986), 317-326.

[2] - Department of Health and Human Services, Centers for Disease Control and Prevention, "Prevention and Control of Influenza Recommendations of Immunization Practices Advisory Committee (ACIP)," Morbidity and Mortality Weekly Report, 41 no. 9 (Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, May 15, 1992), 1-17.

[3] - Department of Health and Human Services, Centers for Disease Control and Prevention, "Pneumococcal and Influenza Vaccination Levels Among Adults Aged 65 Years and Older-United States 1995," Morbidity and Mortality Weekly Report, 46 no. 39 (Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, October 3, 1997), 913-919.

[4] - Department of Health and Human Services, Centers for Disease Control and Prevention, "Pneumococcal and Influenza Vaccination Levels Among Adults Aged 65 Years and Older-United States 1997," Morbidity and Mortality Weekly Report, 47 no. 38 (Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, October 2, 1998), 797-802.

[5] - Nichol, KL, Logfren, RP and Gapinski, J, "Vaccination: Knowledge, Attitudes, and Behavior Among High-Risk Outpatients," Archives of Internal Medicine, 152 no. 5 (1992). 106-110.

[6] - Frank, JW, "Vaccination in the Elderly: 1.Determinates of Acceptance," Journal of the Canadian Medical Association, 132 (February 15, 1985). 371-375.

[7] - Department of Health and Human Services, Centers for Disease Control and Prevention, "Adult Immunization: Knowledge Attitudes, and Practices - Dekalb and Fulton Counties, Georgia 1988," Morbidity and Mortality Weekly Report, 37 no. 43 (Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, November 4, 1988), 395-396.

[8] - Fiebach, NH and Viscoli, CM, "Patient Acceptance of Influenza Vaccination," The American Journal of Medicine, 91 (October 1991): 393-400.