Kaiser-Permanente's Medicare Plus Project: A Successful Medicare Prospective Payment Demonstration

Abstract

The Medicare Plus project of the Oregon Region Kaiser-Permanente Medical Care Program was designed as a model for prospective payment to Increase Health Maintenance Organization (HMO) participation in the Medicare program. The project demonstrated that it is possible to design a prospective payment system that costs the Medicare program less than services purchased in the community from fee-for-service providers; would provide appropriate payment to the HMO; and in addition, creates a “savings” to return to beneficiaries in the form of comprehensive benefits to motivate them to enroll in the HMO.

Medicare Plus was highly successful in recruiting 5,500 new and 1,800 conversion members into the demonstration, through use of a media campaign, a recruitment brochure, and a telephone information center. Members recruited were a representative age and geographic cross section of the senior citizen population in the Portland, Oregon metropolitan area.

Utilization of inpatient services by Medicare Plus members in the first full year (1981) was 1679 days per thousand members and decreased to 1607 in the second full year (1982). New members made an average of eight visits per year to ambulatory care facilities.

To older Americans, the traditional health care system is a vital but bewildering array of medical specialties, hospitals, nursing homes, claim forms and unplanned expenses. No one can erase the physical, psychological and economic problems imposed by advancing age. But the medical care system can move to deal more equitably and effectively with the health problems which place such heavy burdens on older Americans. (Iglehart and Lane)

This paper describes a project which attempts to deal more equitably and effectively with the health problems of older people. The Medicare prospective payment demonstration project (known as Medicare Plus) of the Oregon Region Kaiser-Permanente Medical Care Program (KPMCP) is one of several Medicare experiments funded by the Health Care Financing Administration (HCFA). The project's goal is to increase HMO participation in the Medicare program by designing and implementing a model for prospective payment that would allow Medicare members of an HMO to have prepaid benefits similar to HMO younger members. Such a project should:

1. Cost the Medicare program less than services purchased in the community from fee-for-service providers.

2. Provide appropriate payment to the KPMCP, based on an adjustment of its community rate.

3. Provide a savings to return to beneficiaries as a means of motivating them to enroll in the project and accept the KPMCP as their sole provider of nonemergency medical services.

A basic component of current national health policy is to encourage the development and growth of health maintenance organizations as a cost-effective alternative to the fee-for-service health care delivery system. To that end, it has been proposed that HMOs increase their participation in the Medicare and Medicaid programs. However, to make this attractive to group practice HMOs, it was necessary that Medicare and Medicaid be changed to include HMO operating provisions.

An awareness of the extent to which reimbursement formulas can affect costs and the failure of retrospective cost reimbursement to embody cost consciousness in the delivery of services led to the advocacy of prospective payment under Medicare. This, in turn, led to the development of the Health Care Financing Administration's (HCFA) experimental program in this area,¹ and to the inclusion of prospective payment legislation in the 1982 Tax Equity and Responsibility Act (TEFRA).

Sufficient incentive is needed for Medicare beneficiaries to enroll in HMOs because to do so may mean changing providers and possibly having less freedom of choice of physicians and hospitals.

The Medicare Plus project tests the extent to which this can be accomplished by paying HMOs a meaningful portion of the savings resulting from their efficiency, which then can be passed on to their Medicare members in the form of added benefits, lower rates, or both. This requires HCFA to pay HMOs more than their adjusted community rates for providing Medicare covered services, but will result in HMO members receiving greater benefits than other Medicare beneficiaries. Although this is contrary to the way Medicare has operated previously, it is essential if HMO participation in Medicare is to be increased. Incentives for enrollment in cost-effective systems are a basic requirement for significant delivery system reform. It is economically sound to reward prudent purchasers of health care services.

There are a number of methods for paying HMOs, two principles are essential for the active participation of HMOs on a risk basis:

1. The rate should be determined prospectively on a per capita basis. Both the HMO and the Medicare program should know what the rate will be in advance to allow effective planning and budgeting.

2. The rate should include the savings which an HMO creates through its operational efficiencies when compared to non-HMO costs in the area.

The initial rate setting involves a trade-off between maximum expansion of Medicare membership in HMO's (by including all or most of the savings in the rate) and minimum short-term costs to the Medicare program.

The KPMCP, which is the largest prepaid group practice plan in the United States, has had extensive experience in providing care to Medicare and Medicaid beneficiaries and in participating in the development of Federal and State regulations concerning HMOs. Included among the 4.2 million persons covered in the nine regions of the program are 251,000 Medicare members.

The KPMCP receives payment for Part A (hospitalization) services on a retrospective cost basis using standard Medicare reimbursement rules. Part B payments are based on retrospective cost determination in accordance with the group practice prepayment plan provision of the Medicare Act. KPMCP Medicare members enroll in a supplemental plan which covers the deductible and coinsurance amounts not covered by Medicare and provides selected optional services, such as preventive health services, which Medicare does not cover. Thus, Medicare does not pay the KPMCP a prospectively determined rate, which is the usual way in which an HMO receives payment; nor does the KPMCP have any contracts under Section 1876 of the Act (the Medicare HMO provision).

Although the KPMCP's total Medicare membership is substantial compared to the total size of most HMOs, it is only about 6 percent of the KPMCP's total membership and most Medicare members were members of the Health Plan before they became entitled to Medicare.

The KPMCP has not made substantial efforts to enroll Medicare beneficiaries who are not already members for the following reasons:

1. The benefit or rate incentives to join are inadequate or uncertain.

2. The existing payment provisions (SS1815, 1833, and 1876) are retroactive, which is inconsistent with the KPMCP's basic method of operation.

3. The “lock-in” requirements of Section 1876 are considered difficult, if not impossible, to impose upon existing Medicare beneficiaries who are not currently so restricted.

The KPMCP is able to provide more benefits or lower rates than other insurers because it assures appropriate use of services, especially hospital services. Members use substantially fewer hospital days per thousand persons than comparable fee-for-service populations. In 1978, before this project began, Oregon Region members were hospitalized at the rate of 384 days per 1,000 persons enrolled in the program. This contrasts with the national rate in 1978 of 1,225 days per 1,000 persons. For the population age 65 and over, the rates were 1,630 days per 1,000 for KPMCP members in Oregon contrasted with 4,121 days per 1,000 for the State's aged population (1978).

A similar situation in utilization exists in the Medicare program. KPMCP Medicare members use substantially fewer days than Medicare beneficiaries who obtain services from fee-for-service providers (see Table 1). However, under existing Medicare reimbursement provisions, all savings accrue to the Medicare Trust Fund and not to Medicare beneficiaries. Tables 2 and 3 compare the utilization rates of hospital days and doctor's office visits of members within the Oregon Region who are under age 65 with the rates of members age 65 and over.

Table 1. Hospital Days per 1,000 Persons Age 65 and Over.

	KFHP, NCR	United States	Age/Sex Adjusted KFHP Rates (Assuming U.S. Age, Sex Population Distribution)
Pre-Medicare1	2,322	3,449	2,453
After Medicare2
1967	2,189	3,698	2,9123
1968	2,269	3,990	2,552
1969	2,154	4,048	2,336
1970	2,019	3,904	2,193
1971	1,989	3,835	2,190
1972	1,989	3,835	2,225
1973	1,990	3,853	2,171
1974	1,797	3,963	1,918
1975	1,858	4,003	2,030
1976	1,791	4,121	1,945
1977	1,677	4,156	1,906
1978	1,660	4,184	1,884
1979	1,640	4,182	1,851

¹Data are for the two latest pre-Medicare periods for which such information is available; the year ended June 30, 1963 for KFHP, Northern California, and calendar year 1965 for the US (Source: PHS Publication No. 1000, Series 13, No. 3).

²Utilization data through 1976 for the U.S. general population age 65 and over are from mid-monthly “Hospital Indicators” sections of Hospitals. (Source for 1977, 1978, and 1979: Health United States 1979, 1980, and 1981 issues; DHHS publications No. (PHS) 80-1232, (PHS) 81-1232 and (PHS) 82-1232.) Average population figures used to convert total hospital days to rates per 1,000 were estimates of the resident civilian population as of July 1 of each year. Source: Selected issues of US Department of Commerce Current Population Reports.

³The 1967 hospital day rate is age-adjusted only. Hospital days by male-female distributions are not available.

Table 2. Inpatient Days per 1,000 Health Plan (Oregon Region) Members.

	Younger Than 65 Years	65 Years and Over	Total Health Plan Members
1966	427	1,690	516
19671	388	1,505	473
19681	355	1,313	428
1969	399	1,643	487
1970	371	1,533	449
1971	361	1,572	440
1972	348	1,630	408
1973	329	1,604	405
1974	310	1,679	392
1975	327	1,684	411
1976	309	1,653	396
1977	303	1,707	396
1978	296	1,630	384
1979	300	1,776	399
19802	278	1,651	381
1981	262	1,557	382
1982	273	1,607	401

¹An experimental extended care facility was in operation at Bess Kaiser Hospital and artificially reduced utilization.

²The Medicare Plus project began enrollment in August 1980.

Table 3. Doctor Office Visits per 1,000 Health Plan (Oregon Region) Members.

	Younger Than 65 Years	65 Years and Over1	Total Health Plan Members
1966			3,369
1967	3,279	4,769	3,392
1968	3,192	4,741	3,316
1969	3,104	4,550	3,207
1970	3,280	4,566	3,366
1971	3,307	4,639	3,393
1972	2,981	4,411	3,067
1973	3,015	4,414	3,100
1974	3,136	4,846	3,243
1975	3,043	4,966	3,165
1976	2,995	4,899	3,123
1977	2,915	4,907	3,051
1978	2,761	4,660	2,891
1979	2,567	4,629	2,711
19802	2,546	4,964	2,734
1981	2,559	4,889	2,783
1982	2,555	5,189	2,817

¹Includes under 65 Medicare disabled.

²The Medicare Plus Project began enrollment in August 1980.

Project Design

The specific objectives of the Medicare Plus project, were to develop, implement, and evaluate:

1. A prospective payment system for Medicare members of the Oregon Region of the KPMCP;

2. A system for enrolling the new Medicare members;

3. A service and benefit experiment to test the factors influencing enrollment.

The project design encompasses the essential features of the experimental capitation model outlined by HCFA in the original call for proposals:

1. It is consistent with principles of prepayment.

2. It provides appropriate revenue to the HMO.

3. It is administratively manageable.

4. It provides savings to the Federal government.

5. It promotes the efficient delivery of health services.

6. It has incentives for beneficiaries to enroll.

7. It promotes quality of care.

8. It provides comprehensive health care services.

9. It allows freedom of choice.

Prospective Payment System

Under this experiment, the KPMCP receives payment from HCFA at the beginning of each month for each Medicare Plus member. The payment includes KPMCP's adjusted community rate for Medicare covered services (ACR), and the savings which provide additional benefits. The ACR covers all Medicare A and B services and is adjusted to reflect differences in benefits, utilization rates, and the effective date of the rate and time/complexity factors required to provide services for Medicare enrollees compared to other enrollees of the Health Plan. This ACR is all that KPMCP receives for Medicare covered services. In addition to the ACR, the monthly payment covers all standard Medicare supplemental benefits, plus payment for special new member services. These additional benefits and services are provided from the “savings,” the difference between the ACR and 95 percent of what Medicare calculates it would pay for these beneficiaries in the fee-for-service system (the average adjusted per capita cost or AAPCC).

Each year a rate of payment is calculated for the coming year. This calculation requires the following four steps.

1. Calculate the rates comprising the “AAPCC ratebook”

HCFA's Office of Financial and Actuarial Analysis computes a single rate for each cell of a “ratebook.” There is a cell for each single category of person, characterized by age, sex, county of residence, welfare status, and institutional status. For example, there is a rate for a woman, between age 85 and 89 living in county “A,” not on welfare, but living in an institution. There is a rate for a man, younger than age 65 but disabled, living in county “B,” on welfare, but not living in an institution. The rate in each cell is 95 percent of the projected average per capita cost of non-HMO Medicare beneficiaries in that cell.

2. Forecast population distribution

This step involves forecasting the percentage distribution of aged and disabled Medicare beneficiaries to be enrolled in the next year in each cell of the “overall ratebook.” This was a particular problem for the first year of the project since the distribution to be enrolled was unknown. For the first year (1980), existing KPMCP Medicare membership distributions were used to project age, sex, and county distributions. Welfare membership was projected to be zero and institutionalized membership was estimated to be 0.5 percent. These were conservative estimates since the actual membership was expected to approximate the characteristics of the Medicare beneficiaries of the community, a somewhat older population than the Oregon KPMCP's. The actual characteristics of the Medicare Plus enrolled population were used for projections in subsequent years.

3. Calculate composite monthly capitation rates

This step involves taking a weighted average of the rates to yield a single rate of payment, using the population distributions from Step 2.

4. Recalculate rates of payment retroactively

While the rates calculated in Step 1 are totally prospective, the actual population distribution for each year is used in a final adjustment. If different population characteristics yield a different actual rate of payment, adjustments are made as noted below.

Developing the adjusted community rate (ACR) each year requires the following steps.

1. Compute a program-wide community rate (CR)

The community rate is the per member, per month revenue required to provide prepaid health care services to enrolled members.

2. Disaggregate the CR into specific components

The total forecasted CR is separated into major components of Part A and Part B services and is apportioned to the Medicare cost categories in a manner consistent with current Medicare reimbursement guidelines.

3. Develop adjustment factors

Two types of adjustment factors are necessary to properly reflect the varying cost of providing services to specific populations—volume factors and time and complexity factors. Volume factors reflect different use rates for the various components by the specific population. The time and complexity adjustment takes into account variations in the amount of time and resources necessary to provide a given volume of services to different populations. These are calculated for both hospital and medical services.

Table 4 summarizes the ACR and AAPCC calculations for 1980-1983. In 1980 the difference between these two amounts, that is, the difference between 95 percent of Medicare's average adjusted per capita cost (AAPCC) and the Oregon KPMCP's adjusted community rate (ACR), was $19.71 per month. In 1981 the savings was $19.38, $16.76 in 1982, and $26.76 in 1983. This “savings” is returned to the beneficiary as a “reward” for selecting a more efficient medical care program. Under Medicare Plus, the first priority for use of the savings is to pay for Medicare supplemental coverage.

Table 4. Summary of Payment Calculation/Combined Aged and Disabled.

	1980	1981	1982	1983
95% of Average Adjusted Per Capita Cost (AAPCC)	97.90	113.65	139.65	165.44
Adjusted Community Rate (ACR)	78.19	94.27	122.89	138.68
Savings	19.71	19.38	16.76	26.76
New Member Entry	1.15	1.15	.50	1.00
Benefit Stabilization Fund	3.38	1.10	<2.17>	1.02
Available to Offset Medicare Supplemental Coverage	15.18	17.13	18.43	24.74
Medicare Supplemental Dues	15.18	17.13	23.43	27.74
Required Member Contribution	.00	.00	5.00	3.00

Before this experiment, all Medicare members in the Health Plan were responsible for a monthly premium to cover the cost of Health Plan covered services not included under Medicare and of Medicare deductibles and coinsurance. The Medicare supplemental coverage (M-plan) was developed in order to provide aged KPMCP members the same benefits and access to the program as younger members. In addition to paying M-plan dues, the experiment enhances but does not significantly change the care received by Medicare Plus members and provides some new services. The amount allocated for these new services in the first two years was $1.15 per member, per month.

Any portion of the savings which is not required for current benefit and service packages is retained by HCFA in a benefit stabilization fund (BSF) to smooth out year-to-year variations which are caused by calculating the AAPCC and ACR independently of each other and making annual retroactive adjustments for variances between actual and forecasted demographics. At the end of 1980, the BSF contained $118,616; of this, $77,293 derived from the 1980 payment formula and $41,323 from the retroactive demographic adjustment. This fund grew to $315,000 by year end 1981, and was drawn on in 1982 to moderate the rate increase.

Benefits Experiment

A major purpose of the benefits experiment was to explore the extent to which the KPMCP could attract new Medicare enrollees. These new enrollees would have to give up their previous methods of receiving medical care and agree to receive all their medical services through the KPMCP, except in an emergency. This obviously would be a profound change for some older people, especially if they were satisfied with the medical care they were receiving.

To encourage them to join an HMO, Medicare beneficiaries were offered a variety of health benefits not covered by Medicare. All project enrollees received Medicare supplemental coverage with dues paid from the savings generated by this demonstration. Some Medicare beneficiaries, however, were also offered optional benefits for small additional dues. The experiment was intended to explore which new health benefits or combination of benefits were most effective in recruiting new Medicare members.

Persons applying during the first two months of enrollment were randomly assigned to one of two experiment groups. Half were offered only Medicare supplemental coverage (M-plan) for no monthly cost, while half were offered a choice of the M-plan alone (at no cost) or the M-plan plus the chance to purchase one of three optional benefit packages (see Figure 1).

Randomization was determined by the social security number for new applicants and by the Health Plan identification number for conversion applicants. Families were randomized as a unit based on the first number provided; thus, husband and wife were offered the same coverage options.

Marketing Plan

The marketing plan to recruit 4,000 members began with a two-week media campaign designed to ensure that all Medicare beneficiaries in the service area would be invited to join the project during the six-month open enrollment period. Marketing material also emphasized the need for each individual to weigh the advantages and disadvantages of enrolling based on his/her individual situation and requirements for care.

Television announcements ran in 95 spots (60 or 30 seconds) on all four local commercial stations. They were shown about six times a day during popular viewing times for senior citizens. The television announcement was successful in reaching a very high proportion of the area's senior citizens.

A newspaper announcement including a mail-in coupon and a telephone number appeared 20 times in major local papers and several specialty publications. The media campaign was supplemented before and after by regular contacts with a network of public and private agencies serving the elderly.

The major focus of the marketing plan was to encourage interested Medicare beneficiaries to request information about Medicare Plus. Applications and brochures explaining the program were sent to those who did so. Considerable effort went into developing a recruitment brochure that clearly explained the complexities of the project and outlined the eligibility requirements, the advantages and limitations of joining the program, and the procedures for enrolling.

Care was taken to fully inform potential enrollees of the unique features of the demonstration, such as the need to obtain all services through KPMCP (thus giving up Medicare payment for services performed by other providers). Potential enrollees were informed that the program was subject to change and that they must maintain their Part B coverage. The brochure pages describing the program's limitations are shown in Figure 2 to illustrate how the wording, use of type, and layout contribute to communicating clearly with potential enrollees.

Second Marketing Campaign

The initial target enrollment of 4,000 was assured in July 1980, two months after beginning of marketing. At that time the enrollment limit was raised to 5,500 and a second marketing campaign began to enroll 1,500 additional members by the end of the year. This campaign featured 77 television announcements, a limited number of newspaper announcements, and a news release to about 60 local senior citizen agencies. An inquiry letter was sent to persons who had indicated interest during the first campaign but had not yet applied.

Telephone Center

When enrollment began, a Medicare Plus telephone center was opened in KPMCP administrative offices. Temporary employees staffing the center were given a two-day orientation program and a reference manual so they would provide consistent information to callers. Telephone response was so heavy during the first week of the media campaign that it became necessary to hire and train three additional operators and to add three phone lines to the existing six. A recording device was installed to take messages after working hours. The telephone center remained open for seven months to respond to enrollment requests and to coordinate the enrollment process and new member mailings.

To provide personal assistance to applicants, assistance desks were set up at a number of local senior centers and at KPMCP facilities throughout the metropolitan area.

Conversion Members

The conversion of existing Health Plan members to Medicare Plus was limited in order for Medicare to achieve a net savings on this demonstration. Under the demonstration contract, which is based on what HCFA calculates it would pay for services in the fee-for-service system, HCFA would pay more for an existing Medicare Health Plan member under Medicare Plus than under existing law. Therefore, KPMCP agreed to convert only one Health Plan member for each three new members enrolled.

Brochures and applications were mailed to all 9,000 nongroup Medicare Health Plan members. From the 3,000 who responded, 1,500 were randomly selected and 300 more were put on a waiting list. These 300 additional applicants were accepted when it was assured that new member enrollment would reach 5,500.

A small number of Health Plan members complained about the conversion limitation because they were treated less favorably than new members. Most accepted the explanation that the conversion limitation was necessary to achieve the goal of the demonstration, that is, to change Federal legislation to allow all Medicare beneficiaries the option of receiving medical care on a prepayment basis.

Other Marketing Activities

Due to the success of the television campaign, other marketing activities were very limited. A letter with a tear-off return postcard was sent to 40,000 Health Plan members under age 65 asking them to inform their friends and relatives about Medicare Plus. This was done after the media campaign. Approximately 1 percent responded. During the six-month open enrollment period, a speaker's bureau was maintained and presentations were made to all groups who requested them. In a special effort to reach low-income groups, recruitment material was distributed to all public housing locations and speakers were sent to several public housing meetings. The eight AAA senior citizen centers in the metropolitan area served as information and referral points.

Marketing Campaign Results

The media campaign generated requests for about 15,000 information packets. Those requesting packets were representative of the senior citizens living in the area in terms of county of residence and age (see Table 5). Over two-thirds of the inquiries were made by telephone; most of the remainder came from the mail-in coupons.

Table 5. Marketing Information Requests, by Age of Requestor¹.

Number	Percent	Age
415	7.0%	Under 65
1851	31.3	65-69
1535	25.9	70-74
1135	19.2	75-79
620	10.5	80-84
270	4.5	85-89
72	1.2	90-94
17	0.3	95-99
6	0.1	100 or more
5921	100.0%

¹Includes packets requested through October 31, 1980. Those with unknown age (2692) were excluded.

Approximately 49 percent of the information packets mailed by September 28 resulted in one or more applications being returned for enrollment by October 31. The application response rate was about the same for each of the five-year age categories over age 65 and for urban and rural areas of the five-county area. The response rate was highest (about 52 percent) for telphone requests; mail-in coupons had a response rate of about 39 percent (see Table 6).

Table 6. Percent of Packets Returned by Source of Request.

Percent Returned	Source
51.9%	Telephone or Walk-in
39.2%	Newspaper Coupons
44.3%	Staff Presentation
27.0%	Mail-Out to Under 65 Members
47.6%	TOTAL

(n = 7506 requests)

The marketing campaign was effective in notifying the eligible participants and in attracting people who were likely to enroll. It was also successful in attracting a representative age and geographic cross section of the senior citizen population. This is a significant finding since some people in the Federal government were concerned that only a limited and special subgroup of the aged population would be invited to join the program.

Enrollment Results

The media campaign obtained an impressive response, resulting in 3,500 enrollment request cards submitted to HCFA in June and July, 1980. From these requests, about 2,000 new members were enrolled for August 1 coverage and 1,400 for September 1 coverage. For the remainder of the year, new member enrollment leveled off at 500-600 each month; the target 5,500 membership was reached on January 1, 1981 and a high of 5,886 was reached on March 1, 1981. Applications received after enrollment closed on November 30, 1980 were placed on a waiting list and none of these applications was processed until August 1981 when death and cancellation experience reduced the new membership. Conversion membership reached a high of 1,904 for February 1, 1981 coverage. Table 7 shows year-end membership flow.

Table 7. Medicare Plus Year-End Membership, 1980-1982.

	1980			1981			1982
	New	Conversion	Total	New	Conversion	Total	New	Conversion	Total
Base (only)	2414	800	3214	1581	563	2144	1953	678	2631
Base + SB1	1588	403	1991	2404	714	3118	2447	592	3039
Base + DNT R2	106	14	120	132	35	167	86	24	110
Base + DNT R + SB	997	334	1331	1557	543	2100	1339	462	1801
TOTAL	5105	1551	6656	5674	1855	7529	5825	1756	7581

¹Special Benefits consist of prescription drugs, vision and hearing aids.

²DNT R—Dental Benefit

A total of 655 members died or requested termination during the first coverage year for a termination rate of 7.9 percent. About one-third of these cancellations resulted from death of the member.

Population Characteristics of Enrollees

The population enrolled is somewhat older than the Health Plan's existing over age 65 membership (see Table 8 for comparison of Health Plan and Portland populations). One-sixth of the new members are over 80 years of age and three members are over 100 years of age. The male/female distribution is 40/60 for members aged 65-80 and 35/65 for members over age 80. The proportion of disabled enrollees (4 percent) is similar to the proportion of disabled members in the Health Plan's Medicare population. The enrollment results indicate that a representative age and geographical cross section of the senior citizen population was enrolled. The 5,500 new members represent 4 percent of the eligible population in the five-county enrollment area. This new enrollment brought the KPMCP's proportion up to 17 percent of the total over age 65 population in the market area served by the Health Plan.

Table 8. Medicare Plus Comparative Age Distribution.

Age Group		Medicare Plus	KFHP 65 + (Less Med Plus)	Portland & Salem SMSA B.P.A. Est.
65-69:	Male	13.7	19.6	15.3
	Female	19.5	21.9	19.0
	Total	33.2%	41.5%	34.3%
70-74:	Male	12.2	12.4	11.4
	Female	17.1	14.5	15.4
	Total	29.3%	26.9%	26.8%
75-79:	Male	8.3	6.6	7.3
	Female	12.0	9.4	11.3
	Total	20.3%	16.0%	18.6%
80-84:	Male	4.1	3.4	4.5
	Female	6.9	6.3	8.0
	Total	11.0%	9.7%	12.5%
85+	Male	2.1	1.9	2.6
	Female	4.1	4.0	5.2
	Total	6.2%	5.9%	7.8%
TOTAL		100.0%	100.0%	100.0%

Benefit Experiment Results

There was no statistically significant difference in the proportion of applications returned by those offered basic Medicare Plus at no charge (49 percent) and those offered an additional opportunity to purchase one of three optional benefit packages (47 percent). The experimental randomization was discontinued after two months and, early in 1981, all Medicare Plus members were given a chance to add, drop, or change optional benefits with the result that over 70 percent of members enrolled in one of the three extra packages.

New Member Entry Program

The special services and materials developed for this population were designed to ensure the effective transition of Medicare Plus members into this large, relatively complex program. The new member entry program included a member handbook, a health information form, special reserved appointment procedures, telephone informational tapes, member newsletters, medical office open houses, and, most critically, a Medicare Plus representative. The program was financed during the first year by $1.15 per member per month from the savings.

A key component of the program was the Medicare Plus representative, who played an important role in the development of the new member entry program and in staff orientation. The major functions of the representative were to direct new member orientation, to serve as health care coordinator and ombudsman for project enrollees, and to inform KPMCP operating personnel about the special services, benefits, and circumstances of project enrollees. During the beginning of the project, this office handled at least 1,000 inquiries each month.

A Medicare Plus Member Handbook, designed especially for this population, contained step-by-step information on how to use services (including a contact guide which told the new member what to do to obtain specific services such as medical advice). Information about benefits, doctor appointments, physical examinations, prescription refills, or emergency service was also included. The handbook was written in easy-to-read language and was designed using large print (see Figure 3). A service guide, including physicians' names, a list of facilities and telephone numbers, a map of facilities, and other material, was also produced to assist new members.

A health information form was created to obtain current health status information from members and to identify chronic conditions which might need immediate medical attention. The form was designed using large print with a few simple questions to encourage a high response rate; more than 90 percent were completed and returned. A physician reviewed the forms and the Medicare Plus representative made appointments, if necessary. Appointments were reserved on the schedules of primary care providers for Medicare Plus members who required immediate care or who were anxious to establish a patient-doctor relationship. In addition, project team members designed a protocol for KPMCP pharmacies to make it easier for Medicare Plus members to obtain necessary prescription refills during this transition period.

Recorded telephone tapes gave information similar to that provided in the new member handbook, and telephone numbers for the six tapes were listed in the handbook as well as on a printed card sent to the member's home. Bi-monthly newsletters were published to reinforce information about KPMCP services and Medicare Plus coverage and to provide a means of communication for the Medicare Plus representative.

Assisted by medical office staff, the Medicare Plus representative conducted 15 open houses at various KPMCP facilities. New members were invited to have their questions answered and to learn how to use the medical office of their choice, how to make an appointment, how to get a prescription, and where service departments are located. Approximately 10 percent of new members attended (a considerably higher response than is generally achieved for this type of orientation of Health Plan members).

Utilization

Medicare Plus members used hospital beds at a rate of about 1,677 days per 1,000 members per year during the first 12 months of service (Table 9). This rate is slightly higher than that of other over age 65 members, but it is approximately what was predicted for this population. For comparison, the use rate in this age group in the Portland SMSA in 1978 was 3,142 days per 1,000 people per year, according to data from the Oregon State Health Planning Agency.

Table 9. Medicare Plus Hospital Utilization by Discharge Days.

	Kaiser Foundation Hospitals	Supplemental Beds	Nonemergency Claims	Total Hospital Days	Mean Member Months
19801 days	2148	157	577	2882	19041
Days per 1000 per year	1128	83	303	1514
1981 days	12034	266	299	12599	7505
Days per 1000 per year	1603	35	40	1679
1982 days	11987	40	—0—	12027	7484
Days per 1000 per year	1602	5	—0—	1607

¹August through December 1980

On the other hand, the annualized utilization rate for office visits per 1,000 members is somewhat higher than for other Medicare members and is also somewhat higher than predicted. The data for 1980-1982 are given in Table 10.

Table 10. Medicare Plus Outpatient Utilization.

Visits per 1000 members, per year
	Medicare Plus	Other Medicare	Percent Medicare Plus Higher
1980
Physician	5875	4752	24
Nonphysician	2303	1568	47
TOTAL	8178	6320	29
1981
Physician	5762	4513	26
Nonphysician	2009	1553	28
TOTAL	7771	6063	27
1982
Physician	5780	4914	17
Nonphysician	2010	1521	32
TOTAL	7790	6462	21

The number of visits for this population seems to be relatively stable; therefore, this population may require somewhat more ambulatory care than was predicted (Table 10). This, of course, has implications for both cost and organization of care. For example, early data indicate that estimates of prescription utilization for Medicare Plus members with a prepaid prescription benefit are also too low. This caused a significant increase in the prescription prepayment rate for 1982. Skilled nursing facility utilization was initially lower than predicted and this pattern has continued.

Assessment of the determinants of this utilization pattern has begun and various hypotheses are being offered. One hypothesis that must be considered is that the barriers of the existing Medicare system produced a significant amount of unmet need that has become manifest when these barriers were removed. Another hypothesis is that people who are more likely to select an HMO are those with a higher propensity to use services. It is possible that utilization may be reduced after people become more familiar with the system.

Claims

One of the problems anticipated in the design of the demonstration was ensuring a smooth transition of Medicare beneficiaries from the fee-for-service Medicare system into the KPMCP. One significant aspect of this transition relates to the “lock-in” provision of prospective payment; that is, the requirement that all beneficiaries must receive all nonemergency care in the KPMCP. During the start-up phase, the number of outside claims posed a significant problem. In order to ensure an orderly transition, the Health Plan agreed to pay most claims for out-of-plan use, even though it had a legal responsibility to pay claims only for in- or out-of-area emergencies or for serious illness out-of-area. All first and second Part A claims for covered services received from Medicare Plus members before June 1, 1981 were paid. Most Part B claims during the same period were also paid.

A total of 1,572 claims were paid for all outside services used in 1980. These claims totaled $685,000, of which 74 percent of the dollars and 65 percent of the claims were for nonreferred services. Most of these claims would have been rejected for other Health Plan members but were paid for Medicare Plus members during the first 10 months of the experiment. While 85 percent of the 1980 nonreferred claims were Part B claims, 87 percent of the dollars spent were for Part A.

Almost 90 percent of the outside claims were incurred within the first two months of membership. A higher proportion of older members submitted claims than younger members (Table 11). The 789 members (for whom 1,572 claims were paid) represent almost 12 percent of the total membership. (This table includes members who were referred for outside services, as well as those who submitted claims for non-referred services.) Twenty percent of members for whom outside claims were paid for service in 1980 had terminated by June 1981.

Table 11. Medicare Plus Age Distribution of Members and Members Who Submitted Claims, Year End 1980.

Age Groups	Total Membership	Members with Claims	% of Total
Less than 65	233	38	16.3
65-69	2,130	202	9.5
70-74	1,884	204	10.8
75-79	1,305	169	12.9
80-84	705	103	14.6
85 and over	399	73	18.3
Total	6,656	789	11.9

Several strategies were developed to bring this problem under control. The most important was the implementation in the region of a new position, the Patient Care Coordinator, who was charged with contacting the hospitals of the community and facilitating the transfer of members to KPMCP hospitals.

Claims decreased markedly after 1980. The cost per member, per month for the first six months of 1981 was $10, less than half the amount of the preceding period. As a result of the policy change in June 1981, costs dropped further to $3.22 per member, per month, excluding referred services.

Summary

The Medicare Plus project demonstrated that it is possible to design a workable prospective payment system and that Medicare beneficiaries can be motivated to join an HMO by offering them a premium saving or more benefits than they usually have available. Although outpatient utilization was somewhat higher than predicted, inpatient utilization was near predictions for this population. Initially high claims experience was probably prolonged by a deliberately lenient claims policy, but the problem was controlled by the end of the first 12 months. An annual cancellation rate of approximately 5 percent indicates a high level of member acceptance.

The enrollment of 5,500 new Medicare members into the KPMCP raised the percentage of over age 65 members from 6.8 percent in 1979 to 9.4 percent in 1981. As a result the KPMCP now serves 17 percent of all Medicare beneficiaries in the Portland SMSA. The Health Plan's overall market penetration for 1981 was 19 percent.

These findings indicate the feasibility of public policy encouraging enrollment in HMO's by increasing their participation in the Medicare program. They also demonstrate that increasing Medicare enrollment in HMO's has a potential to help contain Medicare costs and decrease hospital utilization for an increasingly aged population in the United States. The provisions necessary for encouraging more HMOs to compete for Medicare business are now enacted into law in the Tax Equity and Fiscal Responsibility Act of 1982. By year end 1983, the final regulations should be in place to allow all qualified HMO's to enter into prospective payment contracts with HCFA. This first report on the payment, marketing, and enrollment aspects of the KPMCP Medicare Plus demonstration will be followed by a series of research reports of other findings related to utilization, member satisfaction, and provision of new services.