Necessary versus sufficient Data Brief #4 claims data July 2016 KEY FINDINGS An assessment of health care price research implications following the Gobeille v. Liberty Mutual Insurance Co Supreme Court decision The majority of both ERISA and non- ERISA membership was in point of ser- This data brief compares the member- in the ERISA and non-ERISA popula- vice (POS) plans ship and prices of national samples of tions may be driven by prices and/or self-funded and fully insured employ- utilization. This data brief focuses on In 2014, 85% of the ERISA population er-sponsored insurance (ESI) claims comparisons of prices. The analysis and 51% of non-ERISA population were data for the years 2010 through 2014. population was limited to individuals covered by point of service (POS) plans. Differences between these two popu- under 65 years of age enrolled in a Non-ERISA prices were similar to lations has become more relevant large group, commercial health plan ERISA prices on average following the Supreme Court decision with health maintenance organiza- in Gobeille v. Liberty Mutual Insur- tions (HMO), preferred provider or- On average, non-ERISA prices were with- in 5% of ERISA prices for inpatient, out- ance Co, 577 U.S. (2016), which up- ganizations (PPO), or point of service patient, and professional service catego- held an appellate court ruling that (POS) plan types. ries. under the Employee Retirement In- Member demographics were similar come Security Act (ERISA) of 1974, for HMO and POS plans, which com- Figure 2 presents the distribution of age states cannot require self-insured prise the largest portion of the ERISA groups by population and plan type from employers or the administrators of and non-ERISA populations 2014. Between the two populations, the dis- their claims to submit data to all- tribution of membership within plan types payer claims databases (APCDs). In both the ERISA and non-ERISA pop- was similar for HMO and POS plans. Howev- ulations, the largest percentage of er, there were larger percentages of younger Self-insured employers typically fi- membership was in POS plans. In the members (ages 18 – 24 and 25 – 34) en- nance the insurance plan but contract ERISA population, POS plans account- rolled in non-ERISA PPOs compared to with another organization to provide ed for over 80% of membership in ERISA PPOs. services such as provider network every year. In the non-ERISA popula- negotiations and claims processing. tion the total percentage of POS mem- Over the study period, the gender distribu- These organizations are often re- bership increased from 40% in 2010 tion in the ERISA and non-ERISA population ferred to as third party administrators to 51% in 2014 (Figure 1). was similar. On average, membership was (TPAs). In contrast, employers with 49% male/51% female across all plan types fully insured plans purchase insur- The non-ERISA population had nearly (data not shown). ance plans for their employees from the same percentage of the population insurers with administrative services enrolled in HMO plans as POS plans in On average, non-ERISA prices were within included. Insurance plans sold to fully 2010, 38% versus 40%, respectively. 5% of ERISA prices for most plan types and The proportion of non-ERISA HMO medical service categories insured employers are subject to the federal and state insurance regula- enrollment declined from 38% in A price index, controlling for utilization, was tions. Self-funded employers, howev- 2010 to 27% in 2014. ERISA HMO calculated to compare the average non- er, are subject to insurance regula- membership was always substantially ERISA price to the average ERISA price (see tions through ERISA. lower than other plan types but also Data and Methods). An index value of 1.00 decreased from 4% in 2010 to 2% by indicates that the average non-ERISA price is In this analysis, self-funded insurance 2014. The proportion of PPO member- equal to the average ERISA price. Thus, the plan status is used as a proxy for ship was consistent over time in both difference between the price ratio and 1.00 ERISA plans and fully insured status the ERISA and non-ERISA populations. can be interpreted as the percentage differ- as a proxy for non-ERISA. Differences www.healthcostinstitute.org 1 ence in price of the non-ERISA popula- The average price for a specific service analyses, even when ERISA data is not tion relative to the ERISA population. may differ between ERISA and non- available. For example, a price index of 1.05 im- ERISA populations Finally, this study compared prices plies non-ERISA prices were 5% higher It is possible to compare the prices for while attempting to control for utiliza- than ERISA prices. particular services. Although it is not tion differences. The dynamics of the The price indices for each year-plan possible to compare every price for eve- differences between the two popula- type-medical services combination are ry service, as a descriptive example Ta- tions are likely much more complicated. shown in Table 1. Non-ERISA HMO pric- ble 2 reports ratios of non-ERISA to This study does not attempt to account es were, 2%-4% higher from 2010 ERISA prices for the ten most prevalent for differences between ERISA and non- through 2012, but by 2014 were 1%-2% inpatient diagnosis-related group (DRG) ERISA populations due to member pref- lower. Non-ERISA POS prices were con- services in the ERISA population in erences or health plan benefit struc- sistently 1%-5% less than ERISA POS 2014. These ten DRGs account for ap- tures, which may influence plan choice, prices. Non-ERISA PPO prices were the proximately 43% of admissions and utilization of services, and ultimately most different ranging from 3% to 13% over 20% of inpatient spending in both impact the price of services. higher than comparable ERISA prices. populations. Consistent with the inpa- Data and Methods tient average the non-ERISA DRG-level Among the service categories, the HMO prices were most comparable to the To ensure reliable comparisons, the and POS average inpatient prices were ERISA prices for POS plan types. The analyses were limited to the ESI large the most similar differing by 1% or 2% widest variation in prices was among group market, under age 65 who were in most years. The PPO inpatient price PPO prices. covered by health maintenance organi- index increased from 1.06 in 2010 to zations (HMO), preferred provider or- 1.13 in 2012. The index value, however, Policy Implications ganizations (PPO), or point of service did decrease in the remaining years of The differences between the ERISA and (POS) plan type. The resulting average the study period to 1.07 in 2014. non-ERISA populations in plan type dis- annual membership accounts for ap- Within outpatient services, the non- tributions and demographics suggest proximately 22% of the national under ERISA POS prices were 4% less than the that there may also be utilization differ- 65 ESI population each year. ERISA POS prices in every year of the ences between the two populations. To control for differences in utilization study. Although both HMO and PPO non- However, the price indices described between populations, a price index was ERISA prices were generally higher above show how the overall average calculated, which held the service mix than the respective ERISA prices, the prices compare, assuming the same mix for each plan type and time period fixed. average difference for HMO plans was of services are used in both populations. This was accomplished by applying the 1%. Among PPO plans the average dif- Generally, prices for HMO and POS plan ERISA populations’ proportion of ser- ference was 8% and in 2011 non-ERISA types tended to be similar between the vices to the non-ERISA population of the PPO outpatient prices were 10% higher. ERISA and non-ERISA populations. In same plan type and time period and cal- PPO plan types non-ERISA prices ap- Similar to the other service categories, culating a weighted average price for peared to be higher. non-ERISA POS professional services each population. The index value was were lower than ERISA POS prices in the Although the HCCI data includes approx- calculated as the ratio of non-ERISA to study period. Among PPO plans, the non imately 27% of the under 65 ESI popula- ERISA total weighted average prices. -ERISA prices fluctuated between 3% tion in the US in a given year, the data is Indices were calculated for combina- and 6% higher than ERISA prices. For a convenience sample and may not be tions of plan type and major medical HMO plan types, however, there was a representative of the full commercially service category. Medical service catego- steady decline in the price ratio from insured population. As such the results ries were inpatient, outpatient, and pro- 1.04 in 2010 to 0.98 in 2014. There was of this study may not be generalizable to fessional services (e.g., doctors, nurses, also a decline in the price ratio for POS APCDs in all states. However, the results or other non-facility prices). plans over the study period, from 0.98 in suggest that non-ERISA data may be 2010 to 0.95 in 2014. applicable for many policy relevant www.healthcostinstitute.org 2 Figure 1. Membership by plan type 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2010 2010 2011 2011 2012 2012 2013 2013 2014 2014 ERISA Non-ERISA ERISA Non-ERISA ERISA Non-ERISA ERISA Non-ERISA ERISA Non-ERISA HMO POS PPO Source: HCCI, 2016. Figure 2. Membership by age (2014) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% ERISA HMO Non-ERISA HMO ERISA POS Non-ERISA POS ERISA PPO Non-ERISA PPO Ages 0 - 17 Ages 18 - 24 Ages 25 - 34 Ages 35 - 44 Ages 45 - 54 Ages 55-64 Source: HCCI, 2016. www.healthcostinstitute.org 3 Table 1. Price indices by plan type and medical service category HMO plans POS plans PPO plans Inpatient Outpatient Professional Inpatient Outpatient Professional Inpatient Outpatient Professional 2010 1.03 1.02 1.04 0.99 0.96 0.98 1.06 1.08 1.05 2011 1.01 1.01 1.04 0.96 0.96 0.98 1.12 1.10 1.06 2012 1.01 1.03 1.02 0.98 0.96 0.98 1.13 1.07 1.03 2013 0.99 1.01 1.01 0.99 0.96 0.97 1.11 1.07 1.04 2014 0.99 1.00 0.98 0.98 0.96 0.95 1.07 1.08 1.03 Average 1.01 1.01 1.02 0.98 0.96 0.97 1.10 1.08 1.04 Source: HCCI, 2016. Note: The price ratio is calculated as non-ERISA weighted average price divided by ERISA weighted average price. Table 2. Price ratios for top ten DRGs (2014) DRG DRG description HMO POS PPO 775 Vaginal delivery without complicating diagnoses 0.86 0.99 1.17 795 Normal newborn 0.88 1.00 1.10 766 Cesarean section without CC/MCC* 0.85 0.98 1.22 885 Psychoses 0.83 0.94 1.10 470 Major joint replacement or reattachment of lower extremity without MCC* 0.93 1.00 1.10 897 Alcohol/drug abuse or dependence without rehabilitation therapy without MCC* 0.99 0.86 0.99 765 Cesarean section with CC/MCC* 0.85 0.98 1.27 794 Neonate with other significant problems 0.99 1.05 1.09 392 Esophagitis, gastroenteritis, and miscellaneous digestive disorders without MCC* 1.00 1.00 0.89 774 Vaginal delivery complicating diagnoses 0.88 0.98 1.12 Average price ratio of top ten DRGs 0.91 0.98 1.11 Source: HCCI, 2016. Notes: 1. The DRGs listed are the ten most frequent in the overall ERISA population in 2014. The rows are ordered by descending volume. In 2014, the same ten DRGs were also the most frequent with the same rank order among the overall non-ERISA population. 2. The price ratio is calculated as non-ERISA average price divided by ERISA average price. 3. MCC denotes a Major Comorbid or Complicating Condition as defined by CMS. CC denotes a Complicating or Comorbid Condition as de- fined by CMS. Authors Eric Barrette and Katharine McGraves-Lloyd ebarrette@healthcostinstitute.org 571-257-1584 Copyright 2016 Health Care Cost Institute, Inc. Unless explicitly noted, the content of this report is licensed under a Creative Commons Attribution Non-Commercial No Derivatives 4.0 License Health Care Cost Institute, Inc. 1100 G Street NW, Suite 600 Washington, DC 20005 202-803-5200 www.healthcostinstitute.org 4