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 (32) Citing Articles via Google Scholar Google Scholar Articles by Cunningham, P. J. Articles by Sinclair, M. Search for Related Content PubMed PubMed Citation Articles by Cunningham, P. J. Articles by Sinclair, M. Related Collections Managed Care Consumer Issues

Health Tracking

MARKETWATCH

Do Consumers Know How Their Health Plan Works?

Abstract

 
Expanding consumer choice of plans is beneficial only to the extent that consumers make informed choices. Using data from the 1996–97 Community Tracking Study (CTS), this study compares consumers’ responses on four key attributes of their health plan with information provided directly by the plan. Plan attributes relate to choice of providers and access to specialists. Although the accuracy of reporting some individual attributes was fairly high, fewer than one-third of consumers accurately reported all four health plan attributes. In general, consumers tended to over report plan restrictions, especially the need for approval to see specialists.

Previous research has shown that many consumers lack a basic understanding of key managed care concepts and features, although it is not clear whether this lack of knowledge also extends to their own health plan.¹ Failure to understand the delivery-system aspects of their own plan could result in confusion and frustration, which may ultimately affect satisfaction and quality of care.

In this paper we use unique data from the 1996–97 Community Tracking Study (CTS) to compare consumers’ responses to questions about their health plan attributes with information obtained from the actual plans in which they are enrolled. We examine how accurately consumers report on basic plan features related to choice of providers and access to specialists, including (1) whether they must choose from a list or network of providers; (2) whether the plan covers any of the costs for going outside of the network; (3) whether they are required to sign up with a primary care provider; and (4) whether they are required to get a referral or plan approval for specialty care.

Data And Methods

Top Data And Methods Results Discussion NOTES

 
Data. The CTS, a longitudinal study designed to track changes in the health care system and their effects on people, is sponsored by the Robert Wood Johnson Foundation. The CTS household survey was designed to produce nationally representative estimates for the U.S. population as well as for each of the sixty communities in the CTS.²

The survey involved a sample of households using both random-digit dialing as well as a small field sample to represent households without telephones. Families were defined to include all persons eligible for coverage under a typical private health insurance policy. Separate interviews were conducted for each family in the household.

For each family interview, an individual considered to be knowledgeable about the health, health insurance coverage, and health care use of family members was selected as the "family respondent" (usually a parent for families with children). In addition, respondents were asked about the specific attributes of the private insurance plans they and other family members were enrolled in, namely, the plan’s requirements for selecting providers from a network, rules for going outside of the network, and "gatekeeping" requirements (Exhibit 1).³

Methods. The unit of analysis in this study is the family health plan. For each plan reported on by the family respondent (up to three per family), responses to the questions on plan attributes were linked to plans’ responses to these same questions in the follow-back survey. By linking responses in this manner, we estimate the percentage of health plans for which there was agreement between the family and health plan respondents. Plans that had missing values on the attributes for either the family or plan respondent were excluded (less than 3 percent of plans for each attribute). Thus, lack of agreement between family and health plan reports reflects inaccurate reporting on the part of the family respondent, not failure to answer the question.

The sample for this analysis is about 11,600 health plans that included both family and plan respondent information.⁵ These plans were reported on by a total of 10,884 family respondents. About 7 percent of family respondents (751) reported on more than one health plan (mostly two).⁶ The sample includes only private health insurance plans. Public coverage—including Medicare, Medicaid, and Medigap plans—was not included. Almost all respondents were between the ages of eighteen and sixty-four. Only about 100 family respondents—about 1 percent of the total—were age sixty-five or older.

Attributes were divided into network related attributes and gatekeeper attributes (see Exhibit 1). The former include whether plan enrollees were required to select from a network or list of providers, and whether the plan covered any of the costs for going outside the network. The latter include whether enrollees were required to select a primary care provider, as well as whether members needed a referral or approval from the plan to see a specialist.

Results

Top Data And Methods Results Discussion NOTES

 
Frequency of plan attributes. Estimates of the percentage of plans that have restrictions on provider choice and access differ depending on whether the information was reported by the family respondent or the plan respondent. Estimates of the percentage of plans that have a network of providers are lower based on family reports than they are based on plan reports (Exhibit 2), although estimates of the percentage of plans that have no out-of-network coverage are roughly similar. On the other hand, estimates of the percentage of plans that require referral for specialty care based on family respondent reports are more than twice as high as those based on health plan reports, which suggests a very high degree of overreporting by family respondents. The percentage of plans that require patients to sign up with a primary care provider was also higher based on family reports.⁷

Agreement was also higher when the plan had a network and when a primary care provider was required. The one exception was that family respondents tended to be more accurate about reporting that the plan had coverage for out-of-network services (75 percent) than when there was no coverage for these services (61 percent).

Type of health plan versus plan attributes. In addition to the plan attributes, family respondents were asked to identify whether their plan was an HMO, which could then be verified with the information provided by the health plan respondent. The results show that about three-fourths of family respondents accurately reported whether their plan was an HMO, which is a higher rate of agreement than for three of the four attributes (the network attribute being the exception). Unlike the plan attributes, overreporting HMO plan type did not appear to be more of a problem than underreporting it.

It is unknown whether consumers are any more accurate in reporting other plan types, such as fee-for-service (FFS) plans, preferred provider organizations (PPOs), or POS plans. However, it is likely that consumers’ recognition of these terms is lower than for HMOs, especially considering the extensive media attention HMOs have received.⁹

Overall accuracy of reporting. Although the accuracy of reporting on some individual attributes was fairly high, fewer than one-third of respondents (30.3 percent) correctly reported all four attributes.¹⁰ In addition, just over half of respondents (56 percent) accurately reported both network attributes, while just under half (48.8 percent) accurately reported both gatekeeper attributes. However, among the most restrictive plans, agreement between the family respondent and plan respondent on all four restrictions was somewhat higher (50 percent).

Factors affecting family respondents’ accuracy. There were no differences in the accuracy of reporting all four attributes by amount of experience with HMOs, but there were some differences depending on level of plan choice, health status, and level of education. However, the accuracy of reporting all four attributes did not exceed 36 percent for any characteristic we examined.¹¹

Experience with HMOs. One might expect that greater experience with managed care plans—especially HMOs—would lead to more accurate reporting of these attributes. Although agreement rates on all four attributes are slightly higher for respondents with more than one year’s experience with HMOs, these small differences were not statistically significant (Exhibit 4). Differences in agreement rates for network and gatekeeping attributes by level of experience with HMOs were also small and not statistically significant.

One exception to the general trend is that respondents offered only a single non-HMO plan had the second-highest overall rate of agreement (30.4 percent) and had agreement rates on network attributes that were similar to those offered a choice of HMO and non- HMO plans (about 53 percent). It is possible that having a single plan offering with few restrictions is often easier to comprehend than when plans or plan offerings are more complicated, or it may be due to other respondent characteristics that are also associated with relatively higher agreement.

Health status. One might also expect greater accuracy in reporting among respondents with significant health needs, both because plan restrictions may be of greater importance to them and because they are likely to have more direct experiences in knowing how their plan works. Surprisingly, respondents who reported their health as poor were slightly less likely to accurately report all four attributes than were persons who reported their health as excellent, good, or fair, although these small differences were not statistically significant. The less accurate reporting may reflect having fewer choices or other characteristics associated with lower accuracy in reporting (such as socioeconomic status).

Education. As one might expect, respondents with at least a high school education had greater accuracy in reporting all four attributes than did respondents with less than a high school education. These differences appear to be a combination of accuracy of reporting both network and gatekeeper attributes. However, only about one-third of respondents with a college degree accurately reported all four attributes, which is only slightly higher than for respondents who had only completed high school.

Discussion

Top Data And Methods Results Discussion NOTES

 
Enabling consumers to have a greater role in choosing their health plan is an explicit or implicit goal of many policy proposals, including those that would provide tax subsidies to purchase insurance in the nongroup market, create or expand purchasing cooperatives, allow individuals to buy in to the Federal Employees Health Benefits Program (FEHBP), and mandate employers to offer a choice of plans. These proposals may differ in their primary objective, but they all emphasize the role of the consumer in making plan choices and will be much more effective if consumers are able to make informed choices.

Previous research has indicated that consumers do not understand many of the key concepts of managed care; our study results indicate that this lack of understanding also applies to their own health plan. Although the accuracy in reporting some individual attributes was relatively high, only about one-third correctly identified all four attributes related to the use of networks and gatekeepers.

Overreporting of restrictions. In addition, overreporting of plan restrictions on access and physician choice appears to be more of a problem than underreporting is. That is, consumers enrolled in plans that restrict enrollees’ access and choice generally appear to have greater awareness of these restrictions (with the exception of not being able to go out of network), while consumers who are not restricted often believe that they are.

This overreporting may be the result of consumers’ being influenced by media reports on managed care practices—which they incorrectly assume also apply to their own health plan—or negative experiences with their own health care that they incorrectly attribute to plan restrictions.

Implications for satisfaction and access. A lack of understanding of how one’s health plan works also has implications for consumer satisfaction and access to care. Consumers may experience greater barriers to care if they are unaware of plan restrictions regarding provider choice and access, or they may delay getting care because they mistakenly believe that their plan requires them to get authorization or approval to visit a certain doctor or receive certain services. The CTS data show that consumers who mistakenly believed that they needed a referral for specialty care were somewhat more likely to report delays in getting care than were consumers who accurately reported that plan approval was not required (25 percent versus 22 percent overall). On the other hand, consumers who were unaware that their plan required referrals for specialty care were less likely to report delays in getting care than were consumers who accurately reported that plan approval was required (19 percent versus 25 percent). As other findings from the CTS household and follow-back surveys have shown, these findings suggest that people’s perceptions of plan restrictions may be more important in influencing their assessments of care than the actual plan attributes are.¹²

Need for more consumer information. The results suggest the need for more consumer information and education about how health plans work, particularly the managed care" features that are more typical of health plans today. A considerable body of evidence indicates that early efforts to educate consumers have not been very effective and that simply providing consumers with more information about plans will not necessarily improve their decisions.¹³ In particular, more effective methods for explaining plans as care delivery systems are needed, as well as information that allows consumers to evaluate the trade-offs among access, cost, and quality.

Moreover, the delivery-system aspects of health plans may become more complex as a result of recent trends toward more "openaccess" models of care and patients’ rights legislation that mandates greater access to physicians both inside and outside of a plan’s network. In other words, many plans that previously had no coverage for out-of-network services or self-referred specialty care may loosen these restrictions by implementing complex rules for when these restrictions do not apply or for when there is full or partial coverage. These changes are likely to increase the importance of consumer education, since they may greatly increase the complexity of plan rules as well as the need for consumers to be aware of changes to their own health plan and the health plan market in general.

Editor's Notes

 
Peter Cunningham is a senior researcher at the Center for Studying Health System Change in Washington, D.C. Charles Denk is senior program director at Princeton Survey Research Associates in New Jersey. Michael Sinclair is a senior statistician at Mathematica Policy Research, also in Princeton, New Jersey.

The Center for Studying Health System Change is supported in full by the Robert Wood Johnson Foundation. This work was largely completed while Charles Denk was employed at Mathematica Policy Research. The authors thank Joy Grossman, James Reschovsky, Paul Ginsburg, Ann Greiner, and two anonymous reviewers for helpful comments. Bryan Sayer and Leif Karell of Social and Scientific Systems provided excellent programming assistance.

NOTES

Top Data And Methods Results Discussion NOTES

 

1. See, for example, D.E. Nelson et al., "What People Really Know about Their Health Insurance: A Comparison of Information Obtained from Individuals and Their Insurers," American Journal of Public Health 90, no. 6 (2000): 924–928[Abstract/Free Full Text]; S.L. Isaacs, "Consumers’ Information Needs: Results of a National Survey," Health Affairs (Winter 1996): 31–41; and J.H. Hibbard et al., "Can Medicare Beneficiaries Make Informed Choices?" Health Affairs (Nov/Dec 1998): 181–193.
2. R. Strouse et al., Report on Survey Methods for the Community Tracking Study’s 1996–1997 Round One Household Survey, Technical Publication no. 15 (Washington: Center for Studying Health System Change, November 1998).
3. The question on specialty referral in the household survey was originally designed for use in the Agency for Health Care Policy and Research (AHCPR) Consumer Assessment of Health Plans (CAHPS). All other attribute questions asked in the household survey were originally designed for use in the 1996 Medical Expenditure and Panel Survey (MEPS) conducted by the Agency for Healthcare Research and Quality (AHRQ). Some questions were modified prior to their inclusion in the CTS household survey.
4. Since many health plans offer multiple health insurance "products" that differ on the key attributes of interest in this study, it was necessary to identify the specific product in which household survey respondents were enrolled. Health plan respondents identified the specific product based on the name of the plan, the community of residence for the family respondent, and the name of the employer offering the plan, all of which were obtained from family respondents during the household survey interview. Only those "products" that were identified with a high degree of certainty by the health plan respondent were classified as a "match" and included in this study. Plan respondents reported on the attributes for each specific product that was matched.
5. Fifty-three percent of all family-reported plans linked to plans reported on in the follow-back survey. While statistical imputation and weighting procedures were used to correct for nonresponse in the follow-back survey to produce nationally representative estimates of the U.S. population in various types of plans, we use only the 53 percent of family health plans that actually linked directly to a plan obtained in the follow- back survey. Therefore, the estimates in this report were weighted to correct for the clustering of the CTS sample in sixty communities, multiple selection of families within households, and nonresponse on the CTS household survey. However, the weights do not correct for nonresponse in the follow-back survey, nor do they correct for the selection of the family respondent for use in this analysis. Therefore, caution should be used in making inferences about the total U.S. population in this analysis.
6. There was virtually no difference in the accuracy of reporting health plan attributes between family respondents who reported on only one plan and those who reported on several plans.
7. It is noteworthy that there were more plans that required primary care provider sign-up than there were plans that required specialty referral (42.8 percent versus 27.9 percent based on plan reports). While one might expect these percentages to be roughly the same, the discrepancy is largely due to point-of-service (POS) plans that require primary care providers but also allow self-referrals for specialty care.
8. If we assume that three-fourths of those enrolled in POS plans who overstated the need for specialty referral would have correctly answered the question if it had been identical to the question asked of health plans, then the agreement rate for this attribute would increase to 54 percent for plans that did not require a referral, and an overall agreement rate of 64 percent. While higher than the actual agreement rates reported in Exhibit 3, they are still well below agreement rates for other attributes.
9. Isaacs, "Consumers’ Information Needs."
10. Since some of the low level of agreement on all four attributes may be due to differences between the household and plan versions of the specialty referral question, we also calculated overall agreement rates excluding the specialty referral attribute. Overall agreement on the other three attributes combined was 43 percent.
11. Some of the variation in agreement rates by respondent characteristics may be due to correlation with other respondent characteristics or the type of plan they are enrolled in. Therefore, caution should be used in making causal inferences as to the effect of these respondent characteristics on the accuracy of reporting.
12. J. Reschovsky and L. Hargraves, Health Care Perceptions and Experiences: It’s Not Whether You Are in an HMO, It’s Whether You Think You Are, Issue Brief no. 30 (Washington: Center for Studying Health System Change, September 2000).
13. J.S. Lubalin and L.D. Harris-Kojetin, "What Do Consumers Want and Need to Know in Making Health Care Choices?" Medical Care Research and Review 56 (Supplement 1 1999): 67–102.

                      What's this?

This article has been cited by other articles:

				M. McHugh, R. Kang, and R. Hasnain-Wynia Understanding the Safety Net: Inpatient Quality of Care Varies Based on How One Defines Safety-Net Hospitals Med Care Res Rev, October 1, 2009; 66(5): 590 - 605. [Abstract] [PDF]

				K. Kronebusch, M. Schlesinger, and T. Thomas Managed Care Regulation in the States: The Impact on Physicians' Practices and Clinical Autonomy Journal of Health Politics Policy and Law, April 1, 2009; 34(2): 219 - 259. [Abstract] [PDF]

				J. Zwanziger and N. Khan Safety-Net Hospitals Med Care Res Rev, August 1, 2008; 65(4): 478 - 495. [Abstract] [PDF]

				S.-Y. Liang, K. A. Phillips, and J. S. Haas Measuring Managed Care and Its Environment Using National Surveys: A Review and Assessment Med Care Res Rev, December 1, 2006; 63(6_suppl): 9S - 36S. [Abstract] [PDF]

				J. Zwanziger and N. Khan Safety-Net Activities and Hospital Contracting with Managed Care Organizations Med Care Res Rev, December 1, 2006; 63(6_suppl): 90S - 111S. [Abstract] [PDF]

				K. C. Heslin, R. M. Andersen, S. L. Ettner, G. F. Kominski, T. R. Belin, H. Morgenstern, and W. E. Cunningham Do Specialist Self-Referral Insurance Policies Improve Access to HIV-Experienced Physicians as a Regular Source of Care? Med Care Res Rev, October 1, 2005; 62(5): 583 - 600. [Abstract] [PDF]

				J. D. Reschovsky, J. L. Hargraves, and A. F. Smith Consumer Beliefs and Health Plan Performance: It's Not Whether You Are in an HMO but Whether You Think You Are Journal of Health Politics Policy and Law, June 1, 2002; 27(3): 353 - 378. [Abstract] [PDF]

				H. T. Tu and J. D. Reschovsky Assessments of Medical Care by Enrollees in For-Profit and Nonprofit Health Maintenance Organizations N. Engl. J. Med., April 25, 2002; 346(17): 1288 - 1293. [Abstract] [Full Text] [PDF]