Location:
WELFARE - MEDICAL ASSISTANCE (MEDICAID);

OLR Research Report


May 26, 2009

 

2009-R-0216

MEDICAID PRIMARY CARE CASE MANAGEMENT (PCCM)

By: Robin K. Cohen, Principal Analyst

You asked (1) for a general description of Medicaid PCCM, (2) how many states currently use PCCM as a Medicaid service delivery model, (3) how well these programs have worked, and (4) whether any states have decided to abandon this care model.

SUMMARY

Medicaid primary care case management (PCCM) is a model of health care delivery that generally requires a Medicaid enrollee to choose a primary care provider (PCP) who is responsible for coordinating the enrollee's care and is paid a monthly fee for doing so, on top of payment for providing medical services. In general, all medical services are reimbursed on a fee-for-service basis, and care is available 24 hours a day, seven days a week. Federal Medicaid law considers PCCM a form of managed care, but some assert that it is more akin to traditional fee-for-service Medicaid but with care coordination.

In 2006, we reported (2006-R-0550) that as of July 2005, 29 states were using PCCM as a service delivery model for their Medicaid managed care programs, either exclusively or in combination with other service delivery models (e.g., full-risk managed care organizations) and that the number of states using this model had remained relatively constant between 1990 and 2002. The Kaiser Family Foundation currently reports that 30 states were running PCCM programs as of June 30, 2007. Connecticut began its PCCM pilot program earlier this year, bringing the total to at least 31 states.

In the 2006 report, we wrote about 11 states' experiences with PCCM. A recent evaluation of Oklahoma's Medicaid 1115 waiver program, which includes a large PCCM component and has been covering the non-elderly Medicaid population statewide since 2004, suggests that the PCCM model has improved access to primary care, although some of the evaluators' findings suggest that this has not been the case universally.

In 2006 we identified three states that had abandoned PCCM in favor of the more traditional managed care organization (MCO) full-risk model of care. To our knowledge, no other states have ended their PCCM programs since then. But Kaiser also reports that in certain states, PCCM enrollment fell significantly, with a total drop in PCCM enrollment between June 30, 2005 and June 30, 2007 of over one million Medicaid recipients. We have asked the states that have experienced significant PCCM enrollment declines for an explanation and will report their responses under separate cover.

PCCM ENROLLMENT IN THE STATES

Table 1 lists the states that have PCCM programs and their enrollment changes between 2005 and 2007.

Table 1: PCCM ENROLLMENTS — JUNE 30, 2005 TO JUNE 30, 2007 (duplicated)*

State

2005 Enrollment

2007 Enrollment

% Change

2007 PCCM Enrollment as % of Total Medicaid Managed Care Enrollment

Alabama

409,234

376,760

(7.9)

44.0

Alaska

0

0

   

Arizona

0

0

   

Arkansas

499,029

467,713

(6.4)

52.0

California

0

0

   

Colorado

44,570

29,189

(34.5)

6.7

Connecticut [1]

0

0

   

Delaware

0

0

   

Florida

705,665

564,510

(20.0)

25.9

Georgia

872,146

23,943

(97.3)

2.5

Hawaii

0

0

   

Idaho

142,512

151,711

6.5

99.5

Illinois

0

424,400

100

74.7

Indiana

217,535

66,931

(69.2)

11.2

Iowa

134,953

137,985

2.2

32.4

Kansas

86,203

21,969

(74.5)

15.8

Kentucky

331,528

299,579

(9.6)

66.9

Louisiana

761,468

636,429

(16.4)

100

Maine

164,774

171,554

(4.1)

100

Maryland

0

0

   

Table 1 Continued

State

2005 Enrollment

2007 Enrollment

% Change

2007 PCCM Enrollment as % of Total Medicaid Managed Care Enrollment

Massachusetts

276,296

277,403

.4

29.4

Michigan

0

0

   

Minnesota

0

0

   

Mississippi

0

0

   

Missouri

0

0

   

Montana

57,475

44,534

(22.5)

100

Nebraska

37,906

38,703

2.1

16.1

Nevada

0

0

   

New Hampshire

0

0

   

New Jersey

0

0

   

New Mexico

0

0

   

New York

20,002

17,939

(10.3)

.1

North Carolina

797,551

830,773

4.2

92.8

North Dakota

31,921

29,339

(8.1)

100

Ohio

0

0

   

Oklahoma [2]

6,798

10,406

53.1

2.5

Oregon

11,501

8,689

(24.5)

.9

Pennsylvania

272,627

256,391

(6.0)

10.2

Rhode Island

0

0

   

South Carolina

10,892

54,347

399

36.3

South Dakota

76,640

74,071

(3.4)

42.7

Tennessee

0

0

   

Texas

347,101

713,324

106

31.4

Utah

48,220

47,413

(1.7)

11.3

Vermont

87,061

0

   

Virginia

86,017

64,235

(25.3)

14.4

Washington

4,106

3,946

(3.9)

.3

West Virginia

17,831

20,962

17.6

13.4

Wisconsin

0

0

   

Wyoming

0

0

   

Source: Kaiser Family Foundation, Medicaid Enrollment in Managed Care by Plan Type, 2005 and 2007 reports.

* These figures include enrollees receiving both comprehensive and limited benefits and those enrolled in more than one managed care plan.

[1] Connecticut's PCCM program began operating in early 2009. As of May 1, 2009, 162 HUSKY A recipients were enrolled in PCCM, most of who resided in New Haven county.

[2] Until January 1, 2009, Oklahoma's PCCM program used a partial capitation system, and the Kaiser Foundation considered it to be a Prepaid Ambulatory Health Plan (PAHP) rather than traditional PCCM.

OKLAHOMA

In 2009, Mathematica Policy Research, Inc. completed an evaluation of Oklahoma's SoonerCare 1115 waiver program, which is the state's Medicaid managed care program covering almost all of the state's non-elderly Medicaid enrollees. The evaluation is particularly instructive in that the state moved from full-risk MCO-based care to a statewide partial-risk PCCM statewide in 2004, having previously used PCCM only in the state's rural areas where MCO penetration was weak.

History

Like many states, Oklahoma embarked on Medicaid managed care as a way to contain Medicaid costs, beginning the move from fee-for-service to managed care in 1993. At the same time, the legislature decided to create a new executive branch agency, the Oklahoma Health Care Authority (OHCA), by pulling the Medicaid program out of a larger state agency. OHCA's main focus would be health policy, cost containment, and managed care.

An interesting aspect of the PCCM program (SoonerCare Choice), and one that separates it from traditional PCCM programs, is that it included a partial capitation payment to the PCPs. Physicians were paid in advance a monthly payment that covered a fixed set of services, primarily office visits for primary and preventive care; Early and Periodic Screening, Diagnosis, and Treatment (EPSDT) screening; injections; immunizations; and some basic lab and x-ray services. According to the researchers, these services accounted for about 10% of the total Medicaid services a typical Medicaid enrollee received.

The main goal of the capitation was to encourage greater participation by physicians in the Medicaid program, as well as getting them accustomed to this kind of payment arrangement should the state eventually move all enrollees into fully capitated health plans. The capitation was 16% higher than what these services would have cost on a fee-for-service basis. On top of the capitation, providers received bonuses if their child patients met EPSDT screening targets or received recommended immunizations.

Increasing budget pressures in 2002 and 2003 resulted in the state looking for additional Medicaid savings. It eliminated several optional services and imposed cost sharing and benefit limits. It also took an innovative step and developed with actuaries a system for risk-adjusting MCO rates to reflect enrollees' health conditions. And it gave the MCOS more leeway to control their costs by allowing them to impose service limits. At that time, OHCA staff compared the costs of SoonerCare Plus (MCO-based) and SoonerCare Choice and concluded that the state was paying somewhat more in the Plus program. Concurrently, OHCA issued its first report card comparing SoonerCare Choice with Plus and found the two to be similar in terms of performance and quality, leading the state to question the wisdom of paying the MCOs more.

By mid-2003, two of the four MCOs had pulled out of SoonerCare Plus. One of the two remaining plans argued for a rate increase that OHCA thought was too high. During these negotiations, OHCA also told the legislature and the OHCA board that it could perform in-house the Plus program's administrative functions at one-quarter of what MCOs were receiving and with one-quarter of the staff. By the end of 2003, the OHCA board chose to end the Plus program and begin transitioning all its enrollees into the SoonerCare Choice program.

At this juncture, OHCA began enhancing the Choice program to enable it to perform the care management functions that the MCOs previously had done. The legislature authorized nearly 100 additional positions and funds for administrative costs. OHCA hired 28 nurse care managers, most of whom had served as exceptional needs coordinators with the MCOs to coordinate the care of enrollees with special and complex needs.

Most recently, the state has stopped paying the partial capitation rates to cover core services. Instead, it has been moving in the direction of a “medical home” model, which has been the trend in states that started with more traditional PCCM models. This includes (1) a monthly prepaid care coordination payment of about $3 to $13 per member per month depending on the services offered in the practice and patient characteristics; (2) a visit-based fee-for-service (FFS) component which uses the existing FFS payment system; and (3) an expanded performance-based component, which includes factors for EPSDT screening, cervical and breast cancer screenings, physician inpatient admitting and visits, and emergency room (ER) utilization.

Results

Utilization. Since several objectives in the state's waiver program were aimed at improving access to primary care, The Mathematica researchers examined a number of measures to see whether the state had achieved them. It found that between 1997 and 2007, the number of participating PCPs increased from 414 to 595, with the largest increase coming in 2004 when the state ended the MCO-based Plus program. (OHCA reports that in April 2009, just under 1,400 PCPs were participating in SoonerCare Choice.) The program also saw a near doubling in the median number of annual encounters (visits that capitation payments cover) in rural areas from .82 in 1997 to 1.56 in 2007 for adults. For children, the increase was slightly less. (The urban area increases were similar but less reliable given the fact that some members were enrolled in fully capitated plans during this period.)

Also noteworthy was what the researchers saw at the lower tail of the encounter distribution: in 1997, rural providers at the 25th percentile delivered an average of .31 visits per adult member, suggesting that many providers in the beginning were not seeing their assigned members; by 2007, the average had risen to .94 encounters.

Another way to measure the success in access to primary care was to examine preventable hospitalization and ER use rates. For adults, the rate of total preventable hospitalizations decreased significantly between 2003 and 2006 in both urban and rural areas. Similar declines occurred for specific diagnoses of chronic obstructive pulmonary disease, pneumonia, and adult asthma. However, the researchers noted an increase in preventable admissions for hypertension among urban males. In general, these findings were consistent with what was happening at the national level, but in some cases the state did better.

For children, the trends in rates of hospitalization were not statistically significant from 2003 to 2006. The exceptions were a statistically significant decline in asthma-related hospitalizations among male rural enrollees and a significant increase in gastroenteritis-related admissions among female urban enrollees. The asthma change was in line with national trends, while there were no national changes in gastroenteritis from 2000-2004.

The researchers also compared SoonerCare Plus and SoonerCare Choice. While there were some statistically significant findings that favored Choice over Plus, when the researchers added controls, they concluded that Choice had not performed significantly differently than the Plus program with respect to preventable hospitalizations among adults. For children, the only statistically significant change was in the rate of preventable hospitalizations for asthma: an increase occurred in such hospitalizations in the transition to Choice in urban parts of the state.

Finally, the researchers looked at ER utilization between 2004 and 2007, a time during which U.S. rates were increasing. While ER visits by Medicaid enrollees nationwide rose from 80 to 87 per 100 enrollees, OHCA reported a 5% decrease, from 80 ER visits per 1,000 member months to 76 visits.

Likewise, the data showed a shift toward office visits relative to ER visits between 2003 and 2007, although the number fluctuated. The researchers suggested that the decrease in the ER-to-office visit ratio in 2007 could be attributable to OHCA's ER utilization program, which included provider- and beneficiary-focused initiatives. The researchers took a closer look at providers that most frequently sent their patients to ERs to see how OHCA's efforts affected them. Looking at the 95th percentile of providers (those with the highest 5% of ER-to-office ratios), it found that in 2003, these providers reported 2.85 ER visits for every office visit. By 2007, this figure had dropped to 1.26.

Quality. The Mathematica researchers also looked at care outcomes and member satisfaction using both Healthcare Effectiveness Data and Information Set (HEDIS) measures (a set of standardized performance measures for MCOs) and the Consumer Assessment of Healthcare Providers and Systems (CAHPS) surveys, both of which are used by most states to measure quality and satisfaction in their Medicaid managed care programs. The researchers noted the limitations in using CAHPS, which includes measures that seek consumer satisfaction with a health plan as a whole, which might be less relevant in the PCCM model.

The HEDIS national Medicaid mean figures were available for the period 2001 to 2006. The data showed that SoonerCare Choice demonstrated an overall trend of improvement, although in 14 of the 19 measures the program scored below the national mean. These included EPSDT screens for three-to-six year-olds, PCP visits for children over age two, and appropriate asthma medications for individuals aged five to 56. In five of the 19 measures, SoonerCare Choice consistently met or exceeded the national benchmark. These included dental visits for children under age 21, EPSDT visits for children between birth and 15 months, and PCP visits.

The CAHPS results among adults showed small changes in adult satisfaction ratios between 2003 and 2005 that were not statistically significant. When looking only at measures that might compare PCCM and MCOs, the researchers found that 75% of Choice members ranked their overall health care and personal health care providers at seven or higher on a scale of one to 10. Two-thirds said they always or usually got needed care quickly. Yet for each of the measures, the 2005 SoonerCare Choice results were below the 2005 national Medicaid CAHPS benchmark. The SoonerCare Choice enrollees were closest to the national benchmarks in the area of access to care. SooncerCare Choice results were a bit further below the national benchmark in their overall rating of doctors, nurses, and specialists and in their overall rating of their health care.

Costs. The researchers looked at cost growth among SoonerCare members between FYs 1999 and 2005. Oklahoma's average annual growth rate in per-member costs exceeded the national average for all Medicaid eligibility groups except people with disabilities during this period. Compared to other states with PCCM programs, the state had a comparable growth rate for spending on the aged, a lower rate for people with disablities and children, and a much higher rate for adults. These trends were expected given the fact that the state expanded its benefit package and increased physician reimbursement rates to the Medicare level.

Although there was growth within the program, the researchers noted that the absolute level of Medicaid expenditures as a percentage of the state budget in 2006 remained 28% below the national average and 5% below the average among other states with PCCM programs. They also noted that overall, states with MCO and FFS Medicaid systems allotted a relatively higher proportion of expenditures to Medicaid than Oklahoma and other PCCM states did, while states with hybrid MCO-PCCM programs closely tracked the national average.

Mathematica Policy Research, Inc., Sooner Care 1115 Waiver Evaluation: Final Report,” January 2009; OHCA website.

OTHER RESOURCES

Office of Legislative Research, Primary Care Case Management in North Carolina, OLR report 2008-R-0622, ( 2008).

RC:ts/ak