Health Affairs, 25, no. 4 (2006): 1086-1094 doi: 10.1377/hlthaff.25.4.1086 © 2006 by Project HOPE	New Online   House Health Reform Bill   Paying for Reform   Vetting AHIP's Report   HIV/AIDS Costs   Brief: Insurance Reform   HA Blog Top 10

This Article Abstract Reprint (PDF) Online Appendix 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 Web of Science 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 Google Scholar Google Scholar Articles by Hsia, R. Y. Articles by Baker, L. C. Search for Related Content PubMed PubMed Citation Articles by Hsia, R. Y. Articles by Baker, L. C. Related Collections Access To Care Health Reform Hospitals Legal/Regulatory Issues Business Of Health Safety-Net Systems Health Spending Consumer Issues

Health Tracking

MARKETWATCH

Do Mandates Requiring Insurers To Pay For Emergency Care Influence The Use Of The Emergency Department?

Abstract

 
Many states have "prudent layperson" mandates that require health plans to reimburse hospitals for emergency department (ED) care delivered to patients who believe that they have symptoms warranting emergency treatment. Increased, and possibly unnecessary, ED use has often been attributed to these policies. We use data from thirty-five states to study relationships between passage of prudent layperson policies in the late 1990s and ED use among the privately insured. None of the analyses show evidence that the mandates are associated with increased use. We conclude that prudent layperson mandates are not associated with increases in ED visits among privately insured patients.

Responding to these concerns, and to fears that patients’ access to EDs might be overly limited by insurers, a number of states passed "prudent layperson" mandates during the 1990s.² The first was adopted by Maryland in 1993; forty additional states have adopted them since then. These mandates require health maintenance organizations (HMOs) and sometimes other private insurers to reimburse hospitals for ED care delivered to patients who come to the ED with symptoms that a prudent layperson would view as warranting emergency treatment, even if their condition is ultimately determined to be non-urgent.

Some observers fear that these mandates have driven up ED use, and they have been commonly cited as a factor contributing to the recent increase in ED use rates.³ Others have gone further, blaming these mandates for encouraging unnecessary usage and needlessly driving up costs for employers and employees.⁴ Evidence on the impact of existing prudent layperson mandates on use would help interpret the impact of past policies and improve the formulation of new initiatives, but such evidence is almost entirely lacking. Previous work in this area has focused only on the mandates’ effects on insurance coverage, reimbursement patterns, and hospital billing practices, and results are mixed.⁵ One study simultaneously examined eleven patient protection laws, including prudent layperson regulations, but was unable to draw clear conclusions about prudent layperson mandates specifically.⁶ This paper investigates the relationship between passage of prudent layperson mandates and use of ED services.

Study Data And Methods

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 
Data. We used individual-level data from the public use files of the Center for Studying Health Systems Change’s (HSC’s) Community Tracking Study (CTS) household surveys for 1996–97 and 2000–01.⁷ The first was fielded between July 1996 and July 1997, and the second between September 2000 and September 2001. Among other things, the CTS household survey was designed to produce representative estimates of access to care and use of services for the U.S. population. It is primarily a telephone survey, supplemented by in-person interviews of households without telephones to ensure their representation.

For each respondent, the data report the number of ED visits in the twelve months preceding the survey. The data also include a range of other demographic and health status information, which we used to develop additional measures for use as control variables in our analyses. These include age, sex, race, highest education level attained by a member of the household, self-reported health status, family income, and family type. We stratified some analyses based on whether or not respondents reported being in an HMO.

Using secondary sources, we compiled information on the presence of a prudent layperson mandate and its date of passage for each state.⁸ We resolved disagreements between sources by consulting state personnel, American College of Emergency Physicians representatives, or original state legislative documents. Based on these data, we constructed an indicator of whether or not a mandate was in force during the time period over which each respondent reported ED use.

We classified each state with a mandate according to the strength of the mandate using data from a 2001 survey of state personnel responsible for enforcement that asked how likely it was that a violation would be detected and what the most likely enforcement response would be.⁹ We developed two indicators capturing whether or not fines were reported to be a likely enforcement response, as opposed to weaker or no response; and whether or not the chance of detection was reported to be "certain" or "very likely," which we termed "strong detection," as opposed to "likely" or "50–50 chance," which we termed "weak detection."

We also classified each state according to the prevalence of self-insured firms. We used publicly reported data from the Medical Expenditure Panel Survey Institutional Component (MEPS-IC) to determine whether states were above or below the median with respect to the share of firms reporting having one or more self-insured health plans.¹⁰

The total CTS sample contains data on 60,446 people in 1996–97 and 59,725 people in 2000–01. We restricted our attention to people ages 0–64, with private insurance, living in a state that had at least thirty people who met the inclusion criteria. We excluded people with public coverage because access to and payment for ED use by beneficiaries of public programs is governed under other statutes. We also excluded people from six states in which prudent layperson mandates were passed during the time period that either round of the CTS household survey was in the field, because some respondents would have been affected by these mandates, and we could not tell exactly who was and was not. This left us with an analysis sample of 59,475 people, representing thirty-five states.

Statistical analysis. We computed the share of respondents who had an ED visit and the distribution of the number of visits in the two samples according to whether or not respondents lived in states that adopted a mandate between the two surveys. We use chi-square tests to evaluate the statistical significance of differences.

To account for potential confounding, we used regression analyses that model the relationship between the presence of a mandate and the number of ED visits in a twelve-month period. To appropriately handle the fact that we observed counts of visits up to four visits per year, but people with five or more visits were all grouped together, we used censored Poisson regression. The key independent variable indicates respondents in a state with a mandate in effect. Controls include demographic and health status variables and an indicator variable for observations from 2000–01 to account for time trends. These models also include indicator variables for each state, which capture all baseline, fixed characteristics of states including the regulatory climate, health care delivery system attributes, and other population characteristics. With the state controls included, the results can be interpreted as capturing the impact of adoption of a new mandate on within-state changes in rates of ED use, using other states that did not pass mandates at the same time as a control group to capture other trends in use over time. For presentation, we used the regression results to compute the predicted mean number of visits in a hypothetical state with average characteristics, with and without a mandate.

All analyses were weighted and adjusted for the multistage sampling design of the CTS using Stata version 9. An online appendix provides additional information about the sample and variables, as well as complete results from all regressions presented and a number of specification tests.¹¹

Study Results

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 
Respondents to the 2000–01 survey reported higher income levels than those in 1996–97, which might be expected as a result of inflation (Exhibit 1). There are also small, although statistically significant, differences in the age, education, and health status distributions between the two CTS rounds.

Exhibit 2 presents changes in unadjusted ED utilization rates between the two survey periods. In states that adopted mandates during the study period, patterns of ED use changed little over time, and there were no statistically significant differences. There were also no significant changes in use patterns in states that did not change their mandate status during the study period.

The federal Employee Retirement Income Security Act (ERISA) exempts self-insured health plans from state mandates, so the effects of mandates might be stronger in states with fewer self-insured plans. To investigate, we estimated models that allowed for separate effects of mandates in states that had above-and below-median rates of self-insurance. We used the results to derive the predicted change in the number of visits per year per 1,000 population associated with mandate passage in states with high and low self-insurance (Exhibit 4). No relationships are statistically significant. Results in states with both higher and lower rates of self-insurance continue to suggest, if anything, that mandate passage is associated with less ED use.

Discussion And Policy Implications

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 
We evaluated the relationship between ED use and the passage of prudent layperson mandates in the mid- to late 1990s. Anecdotal reports have frequently associated increasing trends in ED use with such mandates, and some opponents have claimed that they foster additional unneeded ED use.¹² Counter to these arguments, we found no evidence in any models that passage of a prudent layperson mandate was associated with increases in ED use. None of the estimated relationships between mandate passage and ED use was statistically significant. Even leaving statistical significance aside, the magnitude of the predicted changes is small, and the point estimates suggest that mandate passage is associated with reductions in ED visits. Nor is there any pattern to suggest that states with lower rates of self-insurance or stronger mandates had more apparent relationships between mandate passage and increased ED use than those with weaker laws (although our ability to measure the strength of the laws was limited).

Possible explanations for lack of relationship. This study does not provide evidence about the reasons that there is no relationship between pr udent layperson mandates and ED use. There are a number of potential explanations, however. We have often encountered the view that decisions about whether or not to seek an ED visit could be insensitive to the likelihood that the insurance company will cover the visit. This could be an important factor, but we are not convinced that it would explain the complete lack of relationship that we observed. Although there are undoubtedly situations in which demand for an ED visit would not respond to price or reimbursement expectations, it does seem sensible that at least some visits would be influenced by the probability that the claim would be denied.

A second set of reasons that the mandates might not have had an effect is that the mandates are limited in important ways. Their reach is limited by ERISA so that people in self-insured plans are not covered, although our analyses did not provide reason to believe that this was an important factor.¹³ Mandates also can be confusing and can vary across states, which can sometimes make it difficult for patients and providers to know exactly who is covered, in what circumstances.¹⁴ Even with concerted effort, we encountered numerous difficulties in locating information about the precise provisions of mandates in some states.

Many mandates also might lack effectiveness because there are no mechanisms for enforcing them. Few states have laid out clear penalties for health plans that violate the mandates, and the avenues that individuals or hospitals could use to seek assistance if their cases are not reimbursed are frequently unclear.¹⁵

Finally, many insurers made generalized changes in ED reimbursement policies during the mid- and late 1990s. These might have been the result of many factors, possibly but not necessarily including the generalized threat of mandate passage. Changes in reimbursement policies might affect utilization rates, but these changes could have taken place in all states at about this time, independent of passage of any specific mandate.

Because this study reports a lack of association, a natural question concerns the amount of statistical power available to detect effects. We believe that our results are not attributable to a lack of statistical power. Leaving significance aside, we found that mandates were associated with only small changes in use, in a direction opposite the common expectation that mandates would drive up use. Although we cannot be entirely sure what would happen in a study with a much larger sample size, if the same results were to be found in a setting where they were statistically significant, our conclusions would be unchanged.

Study limitations. Our results should be interpreted in light of the study’s limitations. The utilization data used in this study were self-reported and might not fully capture actual utilization. We did not investigate the mandates’ longer-term implications beyond the time period studied. Our results reflect the impact of mandates passed in the mid- and late 1990s and might not apply to mandates passed in other years. The data that we used to categorize the strength of mandates were based on self-reports by state personnel and thus could be subject to bias.

A full assessment of prudent layperson mandates would also include additional factors. In addition to use, another important objective of the mandates was to improve EDs’ ability to be reimbursed for care that they were providing. Prudent layperson mandates might have affected reimbursements, even if they did not change use. In fact, according to one way of thinking, a strongly positive outcome for the policies would have been to improve reimbursement prospects for EDs that had become inappropriately restricted, without incurring the potential trade-off of contributing to additional, potentially unneeded, ED use. Assessment of whether changes in reimbursements to EDs were justified or actually occurred is not our focus here, and this question does not appear to be fully resolved in existing literature. Further assessment of these issues seems warranted.

Editor's Notes

 
Renee Hsia (rhsia{at}stanford.edu) is a third-year resident in the Department of Surgery, Division of Emergency Medicine, in the Stanford/Kaiser Emergency Medicine Residency Program in Stanford, California. Jia Chan is a research analyst in the Center for Primary Care and Outcomes Research, Stanford University. Laurence Baker is an associate professor of health research and policy in the Department of Health Research and Policy at Stanford University.

The authors appreciate financial support from the Agency for Healthcare Research and Quality, the Robert Wood Johnson Foundation, and the Emergency Medical Foundation.

NOTES

Top Study Data And Methods Study Results Discussion And Policy... NOTES

 

1. W.W. Fields et al., "The Emergency Medical Treatment and Labor Act as a Federal Health Care Safety Net Program," Academic Emergency Medicine 8, no. 11 (2001): 1064–1069.[Web of Science][Medline]
2. D.A. Hyman, "Consumer Protection in a Managed Care World: Should Consumers Call 911?" Villanova Law Review 43, no. 2 (1998): 409–465; L.A. Johnson, "Coverage Disputes and the Prudent Layperson Standard," Annals of Emergency Medicine 44, no. 4 (2004): 426; and author’s reply, 426–427.[Web of Science][Medline]
3. C.W. Burt and L.F. McCaig, "Trends in Hospital Emergency Department Utilization: United States, 1992–99," Vital and Health Statistics 13, no. 150 (Hyattsville, Md.: National Center for Health Statistics, 2001); G.A. Melnick et al., "Emergency Department Capacity and Access in California, 1990–2001: An Economic Analysis," Health Affairs 23 (2004): w136–w142 (published online 24 March 2004; 10.1377/hlthaff.w4.136); S. Heimoff, "Will Prudent Layperson Please Report to the ER," Managed Care 8, no. 5 (1999): 43–46, 49; and B.J. Bearie et al., "Comparison of Emergency Patients’ and Emergency Physicians’ Assessments of Emergency Medical Conditions: Prudent Layperson Definition," Academic Emergency Medicine 9, no. 5 (2002): 540–541.[CrossRef]
4. L. Page, "Issues Facing American Medicine: Patients Bypassing Primary Doctors for Emergency Care," American Medical News, 11 February 2001, http://www.calacep.org/issues/displayissues.html?ID=193 (accessed 17 May 2006); and B. Cecil, "Effects of Government Mandated Benefits on Health Care Costs," September 2004, http://www.bcbst.com/about/affordability/docs/papers/04-639GovernmentMandatedBenefits.pdf (accessed 8 May 2006).
5. C.B. Irvin and J.M. Fox, "Effect of a State Definition of an ‘Emergency Medical Condition’ Legislation on Medicaid Managed Care Organization Reimbursement," Annals of Emergency Medicine 35, no. 3 (2000): 283–286[CrossRef][Web of Science][Medline]; D.C. Seaberg, J.E. Stimler, and R.L. Wears, "Effect of State Legislation Prohibiting Denial of Emergency Department Patient Claims," Annals of Emergency Medicine 35, no. 3 (2000): 267–271[CrossRef][Web of Science][Medline]; J.E. Tintinalli, "Analysis of Insurance Payment Denials Using the Prudent Layperson Standard," Annals of Emergency Medicine 35, no. 3 (2000): 291–293[CrossRef][Web of Science][Medline]; and M.A. Hall, "The Impact and Enforcement of Prudent Layperson Laws," Annals of Emergency Medicine 43, no. 5 (2004): 558–566.[CrossRef][Web of Science][Medline]
6. F.A. Sloan, J.R. Rattliff, and M.A. Hall, "Impacts of Managed Care Patient Protection Laws on Health Services Utilization and Patient Satisfaction with Care," Health Services Research 40, no. 3 (2005): 647–667.[CrossRef][Web of Science][Medline]
7. Center for Studying Health System Change, "Community Tracking Study 1996–1997 Household Survey Public Use File" (Washington: HSC, 2000); and HSC, "Community Tracking Study 2000–2001 Household Survey Public Use File" (Washington: HSC, 2003).
8. We focused on two sources: M. Stauffer and R.B. Morgan, 2001 State by State Guide to Managed Care Law, ed. D.R. Levy (Frederick, Md.: Panel Publishers, 2000); and K. King, "Issue Paper: Prudent Layperson Status"(Irving, Tex.: American College of Emergency Physicians, 2002), 1–5.
9. F.A. Sloan and M.A. Hall, "Market Failures and the Evolution of State Regulation of Managed Care," Law and Contemporary Problems 64, no. 4 (2002): 169–206.
10. Specifically, we obtained publicly reported estimates of the share of firms with one or more self-insured plans by state and year from the Agency for Healthcare Research and Quality, based on data from the Medical Expenditure Panel Survey, Insurance Component (MEPS-IC); see AHRQ, "Index of Insurance Component Tables (Health Insurance Cost Study), 1996–2003," 15 July 2005, http://www.meps.ahrq.gov/Data_Pub/IC_Tables.htm (accessed 10 April 2006). We then classified each state according to whether it was above or below the median level of the 1996–2001 average share of firms with a self-insured plan.
11. This appendix is available online at http://content.healthaffairs.org/cgi/content/full/25/4/1086/DC1.
12. Burt and McCaig, "Trends in Hospital Emergency Department Utilization"; Melnick et al., "Emergency Department Capacity"; and Cecil, "Effects of Government Mandated Benefits."
13. W.K. Mariner, "What Recourse? Liability for Managed-Care Decisions and the Employee Retirement Income Security Act," New England Journal of Medicine 343, no. 8 (2000): 592–596.[Free Full Text]
14. C.R. Gresenz and D.M. Studdert, "Disputes over Coverage of Emergency Department Services: A Study of Two Health Maintenance Organizations," Annals of Emergency Medicine 43, no. 2 (2004): 155–162[CrossRef][Web of Science][Medline]; R. Shesser et al., "Results of Provider Self-Adjudication Using the Prudent Layperson Standard Compared with the Managed Care Organization’s Emergency Department Claim Review Process," Annals of Emergency Medicine 36, no. 3 (2000): 212–218[CrossRef][Web of Science][Medline]; J.S. Stapczynski, "Is the Prudent Layperson Standard Really a ‘Standard’?" Annals of Emergency Medicine 43, no. 2 (2004): 163–165[CrossRef][Web of Science][Medline]; and J. Li, H.K. Galvin, and S.C. Johnson, "The ‘Prudent Layperson’ Definition of an Emergency Medical Condition," American Journal of Emergency Medicine 20, no. 1 (2002): 10–13.[CrossRef][Web of Science][Medline]
15. Sloan and Hall, "Market Failures."

                      What's this?