This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when Comments are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (7) Citing Articles via Google Scholar Google Scholar Articles by Marquis, M. S. Articles by Long, S. H. Search for Related Content PubMed PubMed Citation Articles by Marquis, M. S. Articles by Long, S. H. Related Collections Insurance - Employer-Based System

DataWatch

Prevalence Of Selected Employer Health Insurance Purchasing Strategies In 1997

Abstract

 
This paper provides information about the nationwide prevalence of selected employer health insurance purchasing strategies. These strategies include raising the share of medical costs borne by employees; the use of quality information in choosing which plans to offer; and direct contracting with provider systems. The data are primarily from the 1997 Robert Wood Johnson Foundation Employer Health Insurance Survey.

Strategies in the purchaser’s tool kit include (1) offering a choice of plans, with an emphasis on managed care; (2) sharing costs and responsibility with employees, by benchmarking employer contributions to the lowest-cost plan and through increased cost-sharing amounts or reduced benefits; (3) measuring plans’ quality of care and using this information in decision making; (4) forming and participating in purchasing coalitions; and (5) bypassing insurers and health plans by contracting directly with provider systems.

Primarily using case-study methods, researchers have selected and interviewed private corporations, purchasing coalitions, and state governments representing the cutting edge of the field. This work provides rich context and institutional detail; it also generates hypotheses. However, the study cases may not represent the practices of typical employer purchasers. Nationally representative data are required to find out if most other employers are following the leaders.

The purpose of this paper is to provide new national estimates of the prevalence of selected employer health care purchasing strategies. In previous work we examined managed competition and pooled purchasing.² Here we focus on three remaining topics: (1) increasing employees’ share of costs by reducing the scope or breadth of benefits; (2) using quality information in choosing which plans to offer; and (3) direct contracting with provider systems. Before turning to our findings, we briefly summarize the previous literature on each of these topics.

Cost sharing.

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 
Citing numerous sources, Robert Kuttner concluded that employers shifted health care cost increases to their workers by incorporating financial inducements to choose employee-only as opposed to family coverage, increasing employee contributions to premiums, capping coverage for special benefits, and increasing deductibles or copayments.³

Quality measures.

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 
Several recent studies have explored rates of employers’ use of quality measures in making purchasing decisions, with quite different results. Judith Hibbard and others found that about half of very large employers in four selected markets use National Committee for Quality Assurance (NCQA) accreditation, Health Plan Employer Data and Information Set (HEDIS) measures, and consumer satisfaction surveys in their purchasing decisions.⁴ They found that employers whose workforce was spread across more states were less likely than were those with more geographic concentration to use HEDIS and satisfaction surveys.

Jon Gabel and colleagues, however, found that only 5–11 percent of large employers used quality information in their purchasing decisions.⁵ Anthony Lo Sasso and others found that only a minority of employers considered quality measures to be very important in their purchasing decisions.⁶ They were considered more important by large employers and members of purchasing alliances, but counter to hypothesis, not by the self-insured.

Direct contracting.

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 
Firms’ directly contracting with provider networks is viewed as a relatively rare practice, nationwide.⁷ However, a recent survey of business coalitions found that about one third negotiated directly with providers for their members, suggesting that direct contracting may be more common among businesses participating in pooled purchasing arrangements.⁸ The best-documented example of direct contracting is the Buyers Health Care Action Group in Minnesota. Tom Bodenheimer and Kip Sullivan suggest that companies are most likely to contract directly alone if they are located in markets with low managed care penetration. Further, a firm is more likely to contract directly if its employees are geographically concentrated in one or a few health care markets.⁹

Data And Methods

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 
Sample. Our data are primarily from the 1997 Robert Wood Johnson Foundation (RWJF) Employer Health Insurance Survey, which collected information from a national sample of 21,545 private business establishments. The sample was drawn from the Dun’s Market Identifiers national census of employment establishments. Establishments in the sixty communities followed by the RWJF’s Center for Studying Health System Change (HSC) Community Tracking Study and in twelve states with significant small-group insurance reforms were oversampled.¹⁰ Within geographic areas, the samples were allocated to strata defined by the number of workers at the establishment. We used sample weights to make estimates for several different units of observation: all private employment establishments in the continental United States that had at least one employee and that offered health insurance, all employees of such establishments, and all enrollees in their employer plans.

Data were collected by computer-assisted telephone interviews conducted with the person or persons most knowledgeable about health benefits and firm and worker characteristics. The response rate to the survey was 60 percent.

For our analysis of increasing employees’ share of costs, we measure trends between 1993 and 1997. To make these estimates we compare data from the 1997 RWJF survey to those from the National Employer Health Insurance Survey (NEHIS). The NEHIS is a national survey of 34,604 private employers, which collected information about health insurance offerings in 1993. Our methods for comparing these survey data are described elsewhere.¹¹ (All differences that we cite in the text are statistically significant at p = .05.)

Concepts and definitions. We used a number of specialized concepts and definitions. We report the actuarial value of plan benefits, measured as the expected plan benefit payments relative to the expected total health expenditures for a standardized population. Also, we examine actuarial values for two subpopulations: health service users whose spending was below the median for all users and those whose spending was above the median.

Our measures of patient cost sharing are the deductible and a coinsurance rate. Deductible amounts for 1993 have been inflated to 1997 dollars for comparison with the RWJF data using the medical care services component of the Consumer Price Index (CPI) for all urban consumers (U.S. city average). Health plans use either dollar copayment amounts or percentage coinsurance rates. Our survey collected whichever approach was used for physician office visits. To convert to a standard percentage scale, we divided the copayments by a standard office visit price of $50.

We measure employee contribution rates for single and family coverage. The NEHIS and RWJF surveys collected information about employer and employee contributions for each plan offered. Some employers offered coverage for special services, such as dental care or prescription drugs, in plans that were separate from the general medical plan. We constructed employee contributions for the package of benefits held by enrollees by adding together the contributions that were paid both for the general medical plan and for any plans covering special services.

We explore whether employers considered quality information when choosing which plans to offer to their employees. The 1997 survey collected these data only for establishments with 500 or more employees. Respondents were asked if their organization considered the following: (1) plan accreditation by a national board or commission, (2) the results of patient satisfaction surveys, and (3) scores on a report card of selected outcomes, such as HEDIS measures. For this analysis we classify an employer as using quality information if it used one or more of these measures.

For employers offering a self-insured plan, respondents were asked, "Does your company or its administrative representative contract directly with particular physician groups or hospitals to provide health care to employees?" We used these data to classify employers as engaging in direct contracting.

Although the sample and analysis unit is the establishment—a physical location of business—for some analyses we categorized establishments according to the size of the firm, which includes employees at all locations nationwide. Insurance decisions typically are made at a regional or national level for firms with several establishments. As a measure of the degree of a firm’s operation in a single health insurance market, we used the number of the firm’s employees working in the same state as the sampled establishment, divided by the number of the firm’s employees nationwide.¹²

We classified employers as self-insured if they offered at least one self-insured plan, based on the respondent’s self-assessment.¹³ Several other classification s were based on self-assessment, including participation in a pooled purchasing arrangement and whether a plan is a health maintenance organization (HMO), a preferred provider organization (PPO), a point-of-service (POS) plan, or a conventional indemnity plan.¹⁴

HMO penetration, measured for the sixty Community Tracking Study sites, is based on tabulations from the employer survey data of the share of employees enrolled in their employer’s plan who were in an HMO. It is defined as "low" if the community is in the bottom quartile, "high" if in the top quartile, and "medium" otherwise.

Study Results

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 
Increasing employees’ share of medical costs. One action an employer could take to raise employees’ costs is to tighten rules governing which of them are eligible for coverage or how long they must wait for benefits. But most discussions of employer strategies to decrease their own costs focus on moves to create better incentives for enrollees to engage in cost-conscious behavior. One strategy is to reduce the value of plans, either by increasing patient cost-sharing amounts or by reducing the number or breadth of covered benefits. An alternative is to increase the employees’ share of premiums. While this might be done across the board, the managed competition approach calls for raising employee contributions most for higher-cost plans to induce enrollment in lower-cost plans.

In contrast to Kuttner’s conclusion, we find little evidence that employees’ share of medical costs increased over the 1993–1997 period (Exhibit 1). Eligibility rules were quite stable, whether we consider the minimum hours of work per week required for eligibility or waiting periods from the date of employment.

Expected insurance payments for spending by health care users whose expenditures were below the median, in contrast, remained unchanged. This stability masks underlying changes in several directions, however. As we have documented elsewhere, enrollees shifted from indemnity plans to HMOs.¹⁵ The latter have lower cost sharing and more generous expected benefits for small health expenditures than other types of plans have. At the same time, cost sharing for small expenditures in HMOs and PPOs increased over the period, as seen by higher coinsurance rates in HMOs and by the fall in the expected insurance reimbursement share for HMO and PPO enrollees with low health care spending. For low-spending HMO enrollees, the reimbursement share fell from 82 percent of spending to 75 percent; for PPO enrollees, from 68 percent to 65 percent. Expected insurance payments for low-spending indemnity plan enrollees, on the other hand, increased. With the vast shift away from indemnity coverage that occurred, this perhaps indicates that only the more generous indemnity plans continued to be chosen.

Benefit breadth also changed. For example, the percentage of enrollees covered for each of four special services—mental health, prescription drugs, and dental and vision care—rose during the period, contributing to higher plan benefit values (data not shown).¹⁶

Finally, employers did not make large net changes in workers’ contribution shares to premiums during the mid-1990s. Moreover, the slight increase observed in Exhibit 1 does not appear to be the result of a rush to adopt the managed competition–style incentives of raising employee contributions for higher-cost plans. That is because few employers practiced managed competition, and others were not moving in large numbers to adopt it.¹⁷

Use of quality information. Nearly 60 percent of large employers said that they used information on quality of care when choosing which health plans to offer their employees; the figure rises to nearly 70 percent of those offering HMOs or POS plans, the types for which quality information is most widely available (Exhibit 2). These estimates are considerably higher than those from most previous studies. The comparison between studies is never exact, however. Other researchers have used slightly different measures of quality information; some have asked only about situations where it was deemed "very important" to decisions; or they have looked at employers in a different size range. In an effort to limit our data to match others’ estimates as closely as possible, especially by comparing employers of similar size and offering a similar range of plan types, our estimates remain much higher than previous ones.

Direct contracting and carve-outs. Although only 5 percent of establishments engaged in direct contracting, they accounted for 19 percent of employees; 15 percent of health plan enrollees were in a direct-contract plan (Exhibit 3). This is because the likelihood of direct contracting rises sharply at larger firms: One-fifth of establishments of firms with 500 or more employees reported that they use direct contracting to arrange for at least one health plan.

We estimate that direct contracting is far more prevalent than is commonly believed. Our respondents might have had other purchasing arrangements in mind when they reported engaging in direct contracting. They may have been referring to carve-out arrangements for specialized benefits (for example, prescription drugs or mental health care), rather than direct contracting for the full benefit package. We asked a separate question about the practice of carving out a specialized benefit. About 60 percent of the establishments reporting direct contracting also reported using carve-outs (Exhibit 3). Unfortunately, our data do not permit us to distinguish between those who do both and those who answered affirmatively to the direct-contracting question (which came first in sequence) when the only direct contracting they practice is through a carve-out.

Discussion

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 
Most of the literature on employer health insurance purchasing strategies has been based on case studies that focus on "pathfinders"; unfortunately, it is difficult to tell how well-traveled these pathways are by all employers. Our national survey data permit estimates of the prevalence of these practices in 1997. In earlier work, we have documented the widespread shift to managed care, the limited adoption of managed competition principles, and modest participation in purchasing pools among both small and large employers. Here we find that increasing employees’ share of medical costs was not widely practiced, at least not during the mid-1990s. A majority of large employers reported using quality information in choosing which health plans to offer, considerably more than has been reported previously. Finally, our respondents reported that they engaged in direct contracting for health benefits far more often than observers had guessed.

We have some cautionary notes about projecting these findings to the future. The mid-1990s was a period of low health cost inflation, and employers were moving rapidly toward managed care plans as their main cost-control effort. Because most covered employees are now enrolled in managed care plans and because premiums now seem to be increasing more rapidly, employers might move to raise employees’ share of costs in the near future. However, a recent survey suggests that as of 2000, employers had shielded employees from premium increases.¹⁸ Our measure of using quality information comprises many types of quality measures and many levels of intensity of use. Although we think that it shows more interest in quality information than earlier studies have done, it does not say how important this information was in the final health plan selection. Finally, our effort to measure direct contracting was, to our knowledge, the first in an employer survey. Direct contracting is a relatively recent practice and, therefore, subject to more ambiguity of definition and name recognition.¹⁹

The health system trends we have identified deserve continued monitoring. In the case of direct contracting, we also encourage further efforts to define and measure the concept. This is especially important in a dynamic health care system where new arrangements among purchasers, insurers, third-party administrators, and providers are constantly being developed.

Editor's Notes

 
The authors are senior economists at RAND’s Washington, D.C., office.

This research was supported by Grant nos. 028561, 031565, and 039498 from the Robert Wood Johnson Foundation (RWJF). Any views expressed herein are solely those of the authors, and no endorsement by the RWJF or RAND is intended or should be inferred. The authors thank Linda Andrews and Roald Euller for their efforts in preparing the survey data files on which this paper is based.

NOTES

Top Cost sharing. Quality measures. Direct contracting. Data And Methods Study Results Discussion NOTES

 

1. See, for example, T. Bodenheimer and K. Sullivan, "How Large Employers Are Shaping the Health Care Marketplace" (First of Two Parts), New England Journal of Medicine (2 April 1998): 1003–1007; Ibid. (Second of Two Parts), New England Journal of Medicine (9 April 1998): 1084–1087; C.C. Lee, D.L. Rogal, and K.B. Kemp, Purchasing Pathfinders: On Their Way, but Still Wrestling with the Realities of Containing Costs and Promoting Quality (Washington: Alpha Center, May 1999); U.S. General Accounting Office, Management Strategies Used by Large Employers to Control Costs, Pub. no. GAO/HEHS-97-71 (Washington: GAO, May 1997); A.T. Lo Sasso et al., "Beyond Cost: ‘Responsible Purchasing’ of Managed Care by Employers," Health Affairs (Nov/Dec 1999): 212–223; and J. Maxwell et al., "Managed Competition in Practice: ‘Value Purchasing’ by Fourteen Employers," Health Affairs (May/June 1998): 216–226.
2. M.S. Marquis and S.H. Long, "Trends unmanaged Care and Managed Competition, 1993–1997,"Health Affairs (Nov/Dec 1999): 75–88; and S.H. Long and M.S. Marquis, "Pooled Purchasing: Who Are the Players?" Health Affairs (July/Aug 1999): 105–111.
3. R. Kuttner, "The American Health Care System: Employer-Sponsored Health Coverage," New England Journal of Medicine (21 January 1999): 248–252.
4. J.H. Hibbard et al., "Choosing a Health Plan: Do Large Employers Use the Data?" Health Affairs (Nov/Dec 1997): 172–180.
5. J.R. Gabel, K.A. Hunt, and K. Hurst, When Employers Choose Health Plans: Do NCQA Accreditation and HEDIS Data Count? (New York: Commonwealth Fund, September 1998).
6. Lo Sasso et al., "Beyond Cost."
7. Bodenheimer and Sullivan "How Large Employers" (Second Part); and J. Christianson, "The Role of Employers in Community Health Care Systems," Health Affairs (July/Aug 1998): 158–164.
8. I. Fraser et al., "The Pursuit of Quality by Business Coalitions: A National Survey," Health Affairs (Nov/Dec 1999): 158–165.
9. Bodenheimer and Sullivan, "How Large Employers" (Second Part).
10. See P. Kemper et al., "The Design of the Community Tracking Study: A Longitudinal Study of Health System Change and Its Effects on People," Inquiry (Summer 1996): 195–206; and 1997Employer Health Insurance Survey: Final Methodology Report (Research Triangle Park, N.C.: Research Triangle Institute, 1998).
11. See S.H. Long and M.S. Marquis, "Stability and Variation in Employment-Based Health Insurance Coverage, 1993–1997," Health Affairs (Nov/Dec 1999): 133–139; and Marquis and Long, "Trends unmanaged Care."
12. Ideally, we would prefer a measure that reflects the number of different markets in which the firm operates, but the measure we used was the best available to us, and we expect that it is correlated with the former.
13. The question was, "Is this a plan that is purchased from an insurance company or HMO, or is it a self-insured plan?"
14. All establishments that offer insurance were asked, "Does your company purchase health insurance through: a health insurance purchasing cooperative or alliance; a business coalition; a multiple employer trust or a multiple employer welfare association; or a trade or professional association or other membership organization?"
15. Marquis and Long, "Trends in managed Care."
16. M.S. Marquis and S.H. Long, "Special Services in Employer-Sponsored Health Plans, 1993 and 1997," Compensation and Working Conditions (Summer 2000): 23–26.
17. Marquis and Long, "Trends in Managed Care."
18. J. Gabel et al., "Job-Based Health Insurance in 2000: Premiums Rise Sharply while Coverage Grows," Health Affairs (Sep/Oct 2000): 144–151.
19. For example, respondents might have had in mind the practice of their third-party administrators negotiating ("contracting for") discounts from providers.

                      What's this?

This article has been cited by other articles:

				M. B. Rosenthal, B. E. Landon, S.-L. T. Normand, R. G. Frank, T. S. Ahmad, and A. M. Epstein Employers' Use of Value-Based Purchasing Strategies JAMA, November 21, 2007; 298(19): 2281 - 2288. [Abstract] [Full Text] [PDF]

				J. R. Gabel, S. H. Long, and M. S. Marquis Employer-Sponsored Insurance: How Much Financial Protection does it Provide? Med Care Res Rev, December 1, 2002; 59(4): 440 - 454. [Abstract] [PDF]

				L. C. Kleinman Conceptual and Technical Issues Regarding the Use of HEDIS and HEDIS-Like Measures in Preferred Provider Organizations Med Care Res Rev, December 1, 2001; 58(4_suppl): 37 - 57. [Abstract] [PDF]

				J. Dubow Commentary Med Care Res Rev, December 1, 2001; 58(4_suppl): 63 - 66. [PDF]