Health Affairs

This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

Perspectives on this article are available by clicking the author's name below:
Donald A. Young and Thomas F. Wildsmith • Thomas R. Hefty

H E A L T H  A F F A I R S : M A R K E T W A T C H 
W E B  E X C L U S I V E 

17 April 2002

Individual Insurance: How Much Financial
Protection Does It Provide?

A $1,000 tax credit should be more than adequate to buy individual
coverage for healthy, young, single males,
but it would not even come close for their middle-aged peers.

Jon Gabel, Kelley Dhont, Heidi Whitmore, and Jeremy Pickreign

ABSTRACT:

This paper examines the comparative financial protection provided by individual and group health insurance. Data sources include two national surveys of employer-based health plans and e-health insurance listings for individual coverage on the World Wide Web. Data on the use and cost of services are from the National Medical Expenditure Survey (NMES), a national household survey of Americans. We estimate that individual insurance pays on average 63 percent of the health care bill, whereas group health insurance pays 75 percent. Deductibles are much higher in individual insurance, and covered benefits are more meager. At 200 percent of poverty, the top 25 percent of health care users with individual coverage would spend 11 percent of their income for out-of-pocket health care expenses, as opposed to 6 percent for persons with group coverage.

For more than five decades group health insurance has been the leading form of private health coverage for Americans under age sixty-five. In 1999 approximately 154 million workers and their dependents, as well as nearly four million early retirees, received their coverage from their employer or former employer.¹ The long-term decline of employer-based insurance has thrust individual insurance, long viewed by the insurance industry as the "residual market," onto center stage.² From 1987 to 1993 the percentage of Americans under age sixty-five covered by employer-based insurance fell from 69.2 percent to 63.5 percent. The vibrant economy of the 1990s added more than twenty million new jobs in six years, but the percentage of the population under age sixty-five with employer coverage increased only slightly, to 65.8 percent in 1999.³ In the spirit of incremental reform, Congress is considering legislation to extend a tax credit to persons who purchase individual health insurance.⁴ The Bush administration now proposes a $1,000 tax credit for individuals and a $3,000 tax credit for families.⁵

What is the structure of this individual insurance market, which the Bush administration and some in Congress would grow through the use of tax credits? Sixteen million Americans purchased individual health insurance in 1999-approximately 7 percent of the total number of nonelderly Americans holding private policies.⁶ Americans from different walks of life buy individual coverage: the self-employed; workers whose employer does not offer health insurance coverage; persons who are not in the workforce; early retirees who are not yet eligible for Medicare; and persons who have exhausted their continuation (Consolidated Omnibus Budget Reconciliation Act, or COBRA) coverage. Individual insurance is most commonly purchased in the Great Plains and Mountain states, where in some states more than 10 percent of the nonelderly population owns individual insurance policies. Nearly 20 percent of self-employed persons and 17 percent of farm workers are covered by an individual policy. Persons ages sixty to sixty-four are nearly three times as likely to be covered by individual health insurance as are those ages twenty to twenty-nine.⁷

Unlike insurance for large employer groups, in most states individual insurance companies review the insurability of each applicant case by case. Not only is a medical history likely to be required, but applicants also are often asked to undergo a physical exam and provide blood, urine, and saliva samples to assess their likely medical claims expenses. Because individual insurance is subject to vigorous medical underwriting, the persons it covers tend to be in good health. Nearly three-fourths of the individually insured report being in very good or excellent health, and only 6 percent report being in fair or poor health.⁸

The costs of administering individual insurance greatly exceed those associated with employer-based health insurance. Fees paid to insurance agents often constitute 10-15 percent of the premium dollar. Whereas administrative expenses consume about 25-40 percent of each premium dollar for individual insurance, they account for about 10 percent of each premium dollar among large employer groups and 15-25 percent in the small-group market.⁹

Recent studies have advanced public knowledge about the workings of the market for individual health insurance. The Journal of Health Politics, Policy and Law published an issue in 2000 dedicated solely to this market.¹⁰ Other studies have increased understanding of this market through the advent of electronic health insurance products found on the World Wide Web. These Web sites provide consumers with detailed descriptions of individual insurance products available at ZIP codes throughout the United States and with quotes on the cost of single and family coverage for persons in excellent and less-than-excellent health.¹¹ Still other studies explore aspects of the individual market, such as underwriting practices and the extent of disability and chronic illness.¹²

Under several simplifying assumptions, this paper examines the degree of financial protection that individual insurance provides to persons who pass its underwriting screens. We compare the financial protection, generally measured as actuarial value-the percentage of the medical bill paid by insurance-for individual health insurance with that for employer-sponsored plans. We estimate expected out-of-pocket spending for persons holding group and individual insurance who are heavy and light users of medical care.

To our knowledge, this paper provides the first analysis of the comparative financial protection afforded by individual and group health insurance since earlier analysis of the 1987 National Medical Expenditure Survey (NMES).¹³ We also believe that the paper provides the first comparative analysis about the range of coverage options, prices, and financial protection provided by indemnity, health maintenance organization (HMO), preferred provider organization (PPO), and point-of- service (POS) plans in the individual market.

Data And Methods

Data. For this study we used three main sources of data (or groups of databases): one database captured data on individual insurance plans in ten states; two national surveys provided data on group health insurance plans from national surveys of employers; and another database provided data on use of services from a national household sample of Americans with either individual or group coverage.

Individual insurance data. In August 2000 our study team at the Health Research and Educational Trust (HRET) used Web sites such as eHealthInsurance.com and QuoteSmith.com to collect data on the provisions of health insurance plans in two ZIP codes in each of ten states. The ten states selected (Colorado, Florida, Minnesota, New Mexico, New York, North Dakota, Oklahoma, Oregon, Vermont, and Washington) were the same states used in an earlier Robert Wood Johnson Foundation (RWJF) survey that examined the impact of state insurance reforms.¹⁴ Within each state, we selected the most urban ZIP code and randomly selected one rural ZIP code. Using a table of random numbers, we randomly selected five individual health plans within each ZIP code. If there was a choice of HMO, PPO, POs, and fee-for-service (FFS) plans, we selected at least one of each type of available plan. Our total sample included 103 different individual insurance products.

For each plan, we collected data on the "standard" monthly premium for a twenty- seven-year-old male, a twenty-seven-year-old female, a fifty-five-year-old male, and a fifty- five-year-old female.¹⁵ We collected as much data as was offered on the Web sites about the benefits and cost sharing for each insurance product. Key variables included deductibles, out-of-pocket limits, copayments and co-insurance, and covered benefits for selected services.

Because no data are available on enrollment in each of the plans in our sample, we estimated plan enrollment as follows. Data from the 1998-2000 Current Population Survey (CPS) provided statistics on the number of persons in the state covered by individual health insurance.¹⁶ We applied national estimates of the percentage of the individually insured rural and metropolitan population to estimate the number of individually insured persons in each of the ten states residing in rural and urban areas. We assumed that each plan in our sample from rural and from urban areas in each state had equal enrollments.

Group health insurance data. One of our data sources was the 1997 RWJF Employer Health Insurance Survey, a national survey of 21,545 private employers with at least one worker. Using computer-assisted telephone interviews, interviewers collected data from the person in the establishment who was most knowledgeable about health benefits. The response rate for the survey was 60 percent. The RWJF survey uses employee-based plan statistical weights, which allows one to generalize about employees' out-of-pocket expenses for each health plan offered by employers included in the sample.

The RWJF survey collected detailed information about benefits for each plan offered by the employer at the establishment site. This included data on deductibles, coinsurance, copayments, and out-of-pocket spending limits for hospital and physician services, as well as cost-sharing requirements for prescription drugs and mental health services.¹⁷ In our analysis we used data matched to the ten states in the individual insurance sample.

To compare premiums and benefits in the individual and group insurance markets for the same year (2000) and in the same ten states, we used the 2000 Henry J. Kaiser Family Foundation/HRET annual survey of employers. This national survey included 1,887 randomly selected public and private employers with three or more workers. From previous analysis we have determined that there was little change in covered benefits and patient cost sharing for employer-based plans over the years 1997-2000.¹⁸ Hence, we believe that actuarial estimates from the 1997 RWJF survey are consistent with Kaiser/HRET data on benefits in 2000.

Data on household expenditures. We used data from the 1987 NMES for persons with employer-based and individual health insurance to generate data on health care spending and use of services in households headed by persons under age sixty-five.¹⁹ NMES is a comprehensive national survey of 15,000 households, including 36,000 individuals, conducted by the Agency for Healthcare Research and Quality (AHRQ). Using the National Health Accounts, Actuarial Research Corporation updated the NMES data to account for inflation. We generated different distributions for families enrolled in HMO, PPO/POs, and indemnity plans to account for different utilization patterns for different types of plans and for different in- and out-of-network cost-sharing requirements for PPO and POs plans.

Calculating actuarial values. Using data from NMES, Actuarial Research Corporation simulated the percentage of incurred medical expenses each health plan would pay for the relevant sample of NMES households. Only households with individual insurance coverage were used to calculate the actuarial values for each individual insurance plan; likewise, only households with group coverage were used to calculate actuarial values for each group plan.

For each individual and group plan, we calculated the following actuarial values: (1) the average percentage of the medical bill paid by all users of health care services; (2) the corresponding percentage of the bill paid by households with below-median spending; (3) the percentage paid by households with spending above the median; and (4) the percentage paid by the top 25 percent of users.

Actuarial values are weighted by expenditures. Hence, a plan with an actuarial value of 0.70 would pay 70 percent of the incurred medical expenses of the individual insurance population. It does not mean that insurance would pay 70 percent of medical expenses of a typical family. In employer-sponsored health plans, 5 percent of the population incurs 52 percent of expenditures.²⁰

We made an adjustment for use of preexisting condition clauses in individual insurance plans. Based on data from the 1996 Medical Expenditure Panel Survey (MEPS), we assumed that approximately 80 percent of individuals would not be subject to preexisting condition clauses because they were covered by a former employer or another insurer.²¹ For those persons subject to such clauses, their claims expenses would likely decrease 20 percent during the first year of coverage.²² Thus, in total, the actuarial value for individual insurance plans was reduced by 4 percent.

Analysis. This study compares the performance of individual and group insurance. In some cases, we contrast the performance of individual FFS, PPO/POs, and HMO plans with their group-coverage counterparts, and we test whether individual statistics on FFS, PPO/POs, and HMO plans are significantly different from the overall individual insurance mean. We have combined PPO and POs plans into one category because it was impossible to determine the plan type from the data on the e-health insurance Web sites.²³ In comparing sample means, we use t-tests to determine if differences in sample means are statistically significant at the .05 alpha level.

Limitations of study methods. The lack of information on enrollment in each individual insurance plan is the main limitation of our study. Unfortunately, not even state insurance departments possess information on the numbers of persons purchasing individual products. The absence of such information compelled us to take a "second-best approach," in which we assumed that within a geographic area, equal numbers of persons purchased each plan. If persons actually purchase higher-cost plans with a higher actuarial value, the study results would underestimate the actuarial value of individual plans and their cost. If people tend to purchase lower- cost plans, actuarial values and costs are overestimated. In contrast, data on group plans reflect actual numbers of persons enrolled in each plan. Given this limitation, readers should regard estimates for the individual insurance market as simulated extrapolations. The paper depicts the actual range of plan coverage, but the exact location of central tendency is uncertain.

A second study limitation is the use of 1987 utilization data (adjusted according to plan- type enrollments) in calculating actuarial values. Since 1987 inpatient hospital use has declined, inpatient mental health services constitute a smaller percentage of spending, and prescription drugs account for a growing portion of spending. Many individual insurance plans do not cover outpatient prescription drugs, whereas inpatient hospital expenses are virtually always covered. On the other hand, some individual plans do not cover mental health services. Therefore, it is difficult to determine if the use of adjusted 1987 data overestimates or underestimates the actuarial value of individual plans relative to group plans.

Study Results

Cost sharing and covered benefits. We estimate that patient cost sharing in individual plans is much greater than it is in group plans, while covered benefits are more meager (Exhibit 1). In-network deductibles average $1,550 in individual insurance plans, compared with $138 in employer-sponsored health plans; out-of-network deductibles average $2,235 and $354, respectively.

Out-of-pocket spending limits are an important feature of insurance for protecting patients from catastrophic financial liabilities when they are stricken with serious illness. In individual FFS plans only 11 percent of beneficiaries have an out-of-pocket limit below $2,000, whereas 56 percent of covered workers enjoy such limits (Exhibit 1).

Group insurance plans had higher levels of coverage for all five covered benefits selected for our analysis. Prescription drugs were covered for 97 percent of persons with employer coverage but only 80 percent of those with individual coverage. Virtually all persons with group health insurance had coverage for inpatient and outpatient mental health services, compared with only 63 percent (inpatient) and 48 percent (outpatient) of persons with individual coverage.

Actuarial value of individual and group plans. Among all users of health care, individual insurance covers on average 63 percent of the bill, and group insurance covers 75 percent (Exhibit 2). The distribution of actuarial values in individual insurance is decidedly different than in group insurance (not shown). For group insurance plans, most plans have actuarial values clustered near the mean of 0.75, meaning that their plan covers 75 percent of the medical bill. Sixty percent of group health plan enrollees are enrolled in plans whose actuarial values are within ten percentage points of this average (from 0.66 to 0.86). For individual insurance plans, 38 percent of enrollees are in plans having actuarial values ranging from 0.42 to 0.82. The five lowest individual plans had actuarial values of 0.11 to 0.18, whereas the five plans with the highest actuarial values were concentrated at 0.90.

Because of high deductibles in individual plans, differences in actuarial values between group and individual insurance are more dramatic for the bottom 50 percent of users. Individual insurance covers just 30 percent of the bill for these users; by contrast, group insurance covers 67 percent of the bill for these users. For the top 25 percent of spenders, individual insurance pays 66 percent of the bill; group insurance pays 85 percent.

Monthly premiums. The average monthly premium for a healthy twenty-seven- year-old male (with no medical conditions) in 2000 was $132 (Exhibit 3), one-third less than the average premium for group coverage in the ten study states. Many plans with very low premiums compete in the individual insurance market. Plans with premiums at the twenty- fifth percentile for a healthy twenty-seven- year-old cost only $68 a month, 60 percent less than the premium cost for group insurance plans at the twenty-fifth percentile of premiums.

By contrast, the average monthly premium in 2000 for a healthy fifty-five-year-old male was $313, about 60 percent more than the average group health insurance premium and 2.4 times as much as the cost of single coverage for a healthy twenty-seven-year-old male. Many low-premium plans appeared in the sample for fifty-five-year-old males, with the premium at the twenty-fifth percentile being $182.

The value of insurance. The overused business term "value" usually connotes quality for the money. To determine whether low- price plans provide much less financial protection, we compared the average actuarial value of the four quartiles of individual insurance, sorted by monthly premiums for fifty-five- year-old males (Exhibit 4). As premiums rise, actuarial value rises monotonically for low users of health care; for all users and for high users, the third quartile has an unexpected higher (but statistically insignificant) actuarial value than the fourth quartile has.²⁴ For the bottom 50 percent of users, the lowest-price quartile of individual insurance plans covers only 6 percent of medical expenses, compared with 47 percent for the highest-price quartile.

Discussion

Through an analysis of four databases and using several assumptions to fix unknown factors, this paper has estimated the cost and financial protection provided by individual health insurance and compared that cost and protection with those of group insurance. Some study findings confirm conventional wisdom. Cost sharing in the form of higher deductibles in individual insurance plans far exceeds that in group insurance. Many individual insurance plans cover neither prescription drugs nor inpatient and outpatient mental health treatment, rendering covered benefits inferior to benefits in group plans. Monthly premiums for a healthy twenty-seven-year-old male were on average one-third less than the average cost for single coverage in the group insurance market, but the monthly premium for a healthy fifty-five-year-old male was 2.4 times as great as the cost for a healthy twenty- seven-year-old male and far exceeded the average figure for group insurance.

The percentage of likely incurred expenses paid by health plans (that is, plans' actuarial value) averages 63 percent in individual insurance and 75 percent in group insurance. For the lower 50 percent of users, individual insurance covers only about 30 percent of the bill, versus 67 percent under group coverage.

Our results are consistent with the assertion that many low-cost policies are available for the uninsured to purchase with a $1,000 tax credit-with some important qualifications. For a healthy twenty-seven-year-old male, one-fourth of plans on the market cost $68 or less a month, or $816 per year. However, for a healthy fifty-five-year-old male (and very few men this age lack a medical history) only one-quarter can get coverage that costs less than $2,184 per year. More importantly, these low-cost plans would pay just 42 percent of incurred health care costs for the individually insured population.

There are no data on premium increases for individual health insurance since 2000. The cost of group insurance rose 11.1 percent in 2001, and analysts project similar or greater increases in 2002.²⁶ Consequently, the cost of individual insurance in 2002 may be 20 percent higher than the figures reported here.

Actuaries and economists view the aim of insurance as protecting families from large and unpredictable expenses. For the top 25 percent of health care users, the average individual insurance policy would leave the insured person liable for $1,829 in out-of-pocket expenses, or 11 percent of income for someone with income at 200 percent of poverty. If a healthy fifty-five-year-old male received a $1,000 tax credit and purchased a policy of average cost, premium expenses plus out-of- pocket expenses would consume 17 percent of that person's income.

The major limitation of our study methods is that data on the number of persons purchasing each product were not available. Hence, readers should view individual insurance figures as simulated extrapolations, or "second- best estimates," and hence treat them with great caution.

Findings from this study are subject to another caveat. They reflect the current individual insurance market-a market that is quite small. Mark Pauly and others believe that with an infusion of tax-credit money, the individual insurance market would grow, and insurers would no longer need to market their products so intensively.²⁷ As more healthy persons entered the market, insurers would not need to underwrite as aggressively, either. This would reduce administrative costs and thereby raise actuarial values and reduce premiums. Others contend that individual insurance can never be efficient without a strong set of market rules.²⁸ Overcoming favorable selection on the part of insurers requires strong market rules. To deter the sickest persons from disproportionately purchasing health insurance-for example, couples who wait to buy individual insurance until they plan on having a baby-an individual mandate is necessary.

There are no prior natural experiments to determine who is correct about the shape of individual insurance in a brave new world of tax credits. What is certain is that in today's world, people who can pass the medical underwriting screen buy a product that gives them far less protection than that given to their peers with group insurance.

The authors thank the Commonwealth Fund for its financial support, and Jennifer Edwards and Cathy Schoen for their helpful comments. We are grateful for the work of Cathy Callahan and Jim Mays of Actuarial Research Corporation in providing estimates of actuarial values of health plans. We thank Susan Marquis and Steve Long of RAND for allowing us to use earlier analyses of the 1997 RAND Survey of Employers. Special thanks are due to Samantha Hawkins and Jennifer Rabideaux for their research assistance.

NOTES

1. P. Fronstin, "Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 2000 Current Population Survey," EBRI Issue Brief no. 224 (Washington: Employee Benefit Research Institute, December 2000).
2. D. Rogal and A. Gauthier, "The Evolution of the Individual Insurance Market," Journal of Health Politics, Policy and Law 25, no. 1 (2000): 3.
3. See P. Fronstin, testimony before the House Ways and Means Subcommittee on Health, 107th Cong., 2d sess., 4 April 2001, 16.
4. The two major parties' inability to reach agreement in late 2001 about the financing of health insurance for the unemployed and uninsured deadlocked the economic stimulus package, a stimulus that both Democrats and Republicans believed was necessary.
5. The Relief, Equity, Access, and Coverage for Health (REACH) Act sponsored by Republicans and Democrats in 2001 proposes income-based tax credits at $1,000 for individuals and $3,000 for families. See L.A. Jackson and S. Trude, "Stand-Alone Health Insurance Tax Credits Aren't Enough," Issue Brief no. 41 (Washington: Center for Studying Health System Change, July 2001).
6. Based on data from the 1999 Current Population Survey, as reported in K. Pollitz, R. Sorian, and K. Thomas, "How Accessible Is Individual Health Insurance for Consumers in Less than Perfect Health?" (Report prepared for the Henry J. Kaiser Family Foundation, 2001).
7. All figures in this paragraph are from U.S. General Accounting Office, Private Health Insurance: Millions Relying on Individual Market Face Cost and Coverage Trade-Offs, Pub. no. GAO/HEHS-97-8 (Washington: GAO, November 1996).
8. Ibid.
9. M.A. Hall, "The Geography of Health Insurance Regulation," Health Affairs (Mar/Apr 2000): 173-184. Using data from the National Association of Insurance Commissioners, Mark Pauly and Alison Percy estimate that administrative costs constitute about 40 percent of the premium dollar. See M. Pauly and A. Percy, "Cost and Performance: A Comparison of the Individual and Group Health Insurance Markets," Journal of Health Politics, Policy and Law 25, no. 1 (2000): 9-26.
10. Journal of Health Politics, Policy and Law 25, no. 1 (2000).
11. eHealthInsurance.com; and QuoteSmith.com.
12. Regarding underwriting practices, see Pollitz et al., "How Accessible Is Individual Health Insurance for Consumers in Less-than-Perfect Health?" Insurance applicants in this study included a twenty-four-year-old woman with hay fever; a thirty-six-year-old man who had his knee repaired ten years ago; a sixty-two-year-old man who smokes, drinks and is overweight; and a thirty-six-year-old man who is HIV-positive. Regarding disability and chronic illness, see E. Simantov, C. Schoen, and S. Bruegman, "Market Failure? Individual Insurance Markets for Older Americans," Health Affairs (July/Aug 2001): 139-149.
13. See P.F. Short and J. Banthin, "New Estimates of the Underinsured Younger than Sixty-five Years," Journal of the American Medical Association 274, no. 16 (1995): 1302-1306.
14. See J.C. Cantor, S.H. Long, and M.S. Marquis, "Private Employment-Based Health Insurance in Ten States," Health Affairs (Summer 1995): 199-211. These states vary greatly in their regulation of individual health insurance, which affects the cost and type of benefits offered. For example, New York and Vermont have community rating, guaranteed issuance, and a fairly comprehensive standard benefit package. Colorado, Florida, and Oklahoma have a laissez-faire approach to regulation.
15. The individual insurance market uses the term "standard" to indicate the rate for healthy individuals with no preexisting conditions.
16. The Henry J. Kaiser Family Foundation Web site, www.statehealthfacts.kff.org, was the source for data on state enrollment.
17. For additional details about the survey, see MS Marquis and S.H. Long, "Recent Trends in Self-Insured Employer Health Plans," Health Affairs (May/June 1999): 161-166.
18. See J. Gabel, S. Long, and MS Marquis, "Employer-Sponsored Health Insurance: How Much Financial Protection Does It Provide for the Healthy and Sick?" Medical Care Research and Review (forthcoming).
19. At the time that Actuarial Research Corporation conducted the analysis in 2000, detailed data from the 1996 Medical Expenditure Panel Survey (MEPS) on the use and cost of services were not available.
20. M.L. Berk and A.C. Monheit, "The Concentration of Health Care Expenditures, Revisited," Health Affairs (Mar/APR 2001): 9-18.
21. MEPS is a two-year national panel survey of about 10,000 households, including 23,000 individuals. The Agency for Healthcare Research and Quality (AHRQ) conducts the survey.
22. Our estimates of the loss of actuarial value due to preexisting condition clauses is conservative. The Health Insurance Portability and Accountability Act (HIPAA) of 1996 allows states to not provide guaranteed issuance for preexisting condition clauses by simply providing access to high-risk pools. Also, persons who formerly held group insurance must have exhausted their COBRA benefits, previously had eighteen months of continuous coverage, and applied for individual coverage within sixty-three days of leaving their employer's coverage.
23. The feature that distinguishes POs from PPO plans is the presence of a primary care gatekeeper. Many plans on the Web sites did not provide information on this matter.
24. The simple correlation coefficient between premium and actuarial value for all users was 0.70.
25. Two-thirds of the uninsured earn 200 percent of poverty or less. See S. Glied, Challenges and Options for Increasing the Number of Americans with Health Insurance (December 2000), 7, www.cmwf.org/programs/insurance/glied_workable_415.pdf (5 April 2002), based on March 2000 Current Population Survey (CPS) data.
26. See J. Gabel et al., "Job-Based Health Insurance in 2001: Inflation Hits Double Digits, Managed Care Retreats," Health Affairs (Sep/Oct 2001): 180-186.
27. Pauly and Percy, "Cost and Performance," 25.
28. Even Pauly and Percy observe, "The individual market will never be as low in cost as the large group market, and shopping for individual insurance will continue to be a difficult, anxious task, especially for above average risks." Ibid.

Jon Gabel is vice-president, health system studies, at the Health Research and Educational Trust (HRET) in Washington, D.C. Kelley Dhont is a research associate, Heidi Whitmore is a senior research associate (on detail to the Colorado Hospital Association in Greenwood Village, Colorado), and Jeremy Pickreign is a statistician, all at the HRET.

©2002 Project HOPE–The People-to-People Health Foundation, Inc.