Etti G. Baranoff, PhD; Dalit Baranoff, PhD; Thomas W. Sager, PhD; and Bo Shi, PhD
Reform brought about as a result of the Affordable Care Act (ACA) in 2010 represents substantial change for the US healthcare system, as described by Salber and Selecky in their 2014 article.1
In their conclusion, these authors raised a number of questions about the ongoing effects of the ACA. It is our hope that our study can respond to some of their queries, as we provide trends on numbers of insured individuals and healthcare utilizations, using annual statement data from the US health insurance industry from 2006 to 2014 (Table
). These trends include insured plans for group insurance provided by employers (excluding self-insureds), federal employee coverage, individual plans, Medicare, and Medicaid. We have data for the membership (number of insureds) in each subpopulation, and utilization factors of encounters, admission to hospitals, and days in the hospital. Before presenting the trends data, we provide schematic models (Figure 1
[A and B]) that illustrate the primary changes that the ACA was intended to bring about in the context of important exogenous macro factors. Finally, we present and discuss the trends, which are computed from the health insurers’ own data.
The literature on the impact of the ACA is vast and rich. However, it tends to focus on specific utilization items or specific subpopulations rather than on comprehensive utilization factors of the whole population served by US health insurers in insured plans. Therefore, we offer a more comprehensive view in this paper, which depicts the healthcare system in the United States before and after the ACA and sketches the expected impacts of the ACA, the impact of the financial crisis, and demographic trends on each subpopulation. Lastly, we provide trends in membership and utilizations using encounters, admissions, and days in the hospital for the period 2006 to 2014. Trends for group employers’ insured plans, with reductions in membership and utilization, are strikingly different from trends of other subpopulations. The Medicaid subpopulation showed the most dramatic overall growth; the Medicare subpopulation also grew substantially, as did individual coverage, although both to a lesser extent. Growth accelerated with the introduction of state insurance exchanges.
The ACA Model and Expected Changes
The ACA changed the US healthcare system from the model shown in Figure 1A, before the ACA, to the model shown in Figure 1B, after the ACA was enacted. These figures illuminate the reforms that took place in the US healthcare system from the system stakeholders’ point-of-view, and are broad in nature. The details of these changes are enumerated below.
The Major Elements of the US Healthcare System Before the ACA
As shown in Figure 1A, the health insurance industry was a centerpiece of the healthcare system before the ACA. Health insurers had dual roles as: 1) financial intermediaries that move healthcare funds from consumers (ie, employers, individuals, and the government) to providers; and as 2) managers of healthcare via networks of providers and a system of negotiated discounts and utilization reviews. Federal regulation was minimal, except for Medicare and Medicaid (which were jointly regulated by states and the federal government). For the insurance industry, regulation was exercised by state insurance commissions, which supervise for solvency and conduct. Before the ACA, the most striking element was the large group of consumers without health insurance coverage. Being outside of the insurance system, they had no access to the discounts enjoyed by those within the managed care system. With limited access to (affordable) healthcare, the uninsured often resorted to emergency departments for their medical care.
Major Changes in the US Healthcare System Introduced by the ACA
Signed into law in March 2010, the ACA aimed to provide universal healthcare coverage to all Americans. Collaterally, the role of federal regulation increased; however, insurers retained their dual roles as financial intermediaries between consumers and providers and as managers of the utilization of the medical system. The ACA created new state insurance exchanges and high-risk pools for consumers not traditionally covered; insurers entered into these new market niches, as well. The following points summarize key components of the US healthcare system under healthcare reform:
The insurance industry retains its position as the financial intermediator. It transfers healthcare funds from the consumers, insureds, employers, and the government to the healthcare providers; manages healthcare through the managed care system of discounts, networks of providers, and utilization reviews; and its insurance policies extend and restrict coverage and payments through a complex system of deductibles, coinsurance, co-pays, and annual maximum out-of-pocket expenses. However, the ACA has imposed the following substantial new restrictions on health insurers, as also enumerated by Salber and Selecky1
and expanded here:
Employers’ group health insurance.
Profits that result from medical loss ratios less than 85% for health coverage sold to large employers are subject to being returned to policyholders. Small and large insured employers receive a refund for plans with loss ratios less than 80% or 85%, respectively.
Insurers may no longer impose lifetime limits on the cost of medical expenses for the insured.
Insurers are required to include adult dependent children through age 26 on their parents’ health insurance policies.
Insurers may no longer deny coverage on the basis of preexisting conditions of insureds.
All previously underinsured/uninsured individuals are required to obtain health insurance coverage. However, one can still legally escape buying health insurance by paying an annual fine (a “tax,” per the US Supreme Court).
Employers historically have offered, and continue to offer, group healthcare benefits. However, the ACA has imposed significant new regulations on employers who offer group coverage. The following are the most substantial of these new restrictions:
Added population to group health insurance.
Fully insured and self-funded employer group plans are subject to a number of new taxes and fees. For example, plans that exceed certain high-dollar thresholds will be taxed 40% of the exceedance. Employers prefer to avoid this cost and are taking drastic measures to remain below these thresholds.
Employer mandate of “pay or play” (enacted in 2015 as a shared responsibility provision in the ACA) requires employers to offer all eligible employees minimum and affordable coverage. For this requirement, annual limits and lifetime maximums must be removed. In addition, employers are required to maintain premium contribution levels that do not allow an employee’s contributions to exceed 9.5% of the employee’s income. Furthermore, employers are required to offer all eligible employees a minimum level of coverage that is equivalent to a bronze-level plan on the federal health exchange. Under the ACA, part-time employees are mandated to be covered.
Literature on the value of cost sharing is limited to a specific utilization or subpopulation. Thus, the hypothesis that cost sharing may be beneficial for health insurance is not necessarily proven. One example of a favorable view of cost sharing is the recent study by Robinson et al,2 who studied the impact of cost-sharing design on drugs at the California Health Insurance Exchange. They found that “patients can be shielded from the most onerous cost-sharing burdens for specialty drugs while keeping premiums affordable for the entire enrolled population.”
As noted above, children can be covered through age 26, regardless of status (eg, student, marital, work, other coverage) and there are no preexisting conditions. In addition, new employees do not have to wait for coverage past 90 days from date of hire. Due to these changes, and other factors described in eAppendix 1
(eAppendices available at www.ajmc.com
), one might anticipate growth in the insured members under group health insurance. But instead, we find declining numbers.
Under the ACA, all Americans are required to have health insurance (or pay a penalty). Those who meet income limits are eligible for subsidies or coverage under Medicaid. Among the numerous articles on ACA Medicaid expansion one that is especially relevant to our study is, “Top ACA News: Medicaid Expansion, Plan Satisfaction, and More.”3
Here, Joszt notes large growth in Medicaid members and utilization, and observes that there is a lot to uncover for “Republican presidential hopefuls [who] may all be ready to repeal and replace the Affordable Care Act,” as “the health law continues to make big changes to the US healthcare system.” We also found large growth in Medicaid members and utilization (see the next main section on “Trends in Membership and Utilization” for more information).
Coverage expansion in the exchanges and health marketplaces.
Coverage is provided to individuals who were unable to receive coverage because of preexisting conditions (ie, those who did not meet the Medicaid income benchmark and those who did not receive coverage under their employers).
Increases in longevity of the US population and the large contingent of baby boomers entering retirement suggests continuous growth in the Medicare population and utilization independent of the ACA. Furthermore, it appears that specific elements characterizing this subpopulation in relation to health utilization is increasing use further. Brandt et al4
studied the “Methodological Effects on the Measurement of Repeat Hospitalizations” of the Medicare population and concluded that, “Results suggest measurement of readmissions incentivizes inefficient behavior.” Our results support this trend.
The scope of federal regulation of the health insurance system has expanded substantially. At the same time, the scope of state regulation has not diminished.
In summary, the natural expectation is that the trends in health insurance membership and healthcare utilization for each of the 5 subpopulations would be influenced by the 2008 financial crisis, the growth in the Medicare population, expansion of medical technologies, and ACA implementation. For each subpopulation, one naturally expects growth—except, possibly, for the group health insurance provided by employers. Major negatives for the possibility of growth in employer-provided insurance are the reduction in employment that followed the 2008 financial crisis and the additional disincentivizing requirements of the ACA after 2010. However, employment began to grow again about 2 years after the crisis (eAppendix 2
). A priori, the net effect is uncertain; employer group insurance members and utilization continued a secular decline that began even before the crisis.
Trends in Membership and Utilization as Extracted From US Health Insurers’ Annual Statements Data
The changes in the healthcare landscape that the ACA introduced, together with other macro-level factors outlined in the prior section, set our expectations for the trends in utilization among the covered subpopulations. The 2008 financial crisis, the aging of the US population, and the beginning of the exchanges as alternatives for sources of coverage are among the factors, which may work in conjunction with, or in opposition to, ACA-only factors to affect trends.
In this section, we examine the actual trends using US health insurers’ own annual statement data that is filed with the National Association of Insurance Commissioners (NAIC). We disaggregated and tabulated the data by covered subpopulation and type of utilization. The data include all insurers that report to the NAIC as health insurers; HMOs are included, but not self-insured plans that are administered by the health insurers as third-party administrators. The data also do not include insurers that report to the NAIC as life insurers, some of whom provide health insurance among their lines. Following are the step-by-step examinations we conducted.
Data Mining Methodology
First, for each year from 2006 to 2014, we sorted every insurer’s members (insureds) into the following subpopulations: working individuals in group comprehensive coverage, insured under individual comprehensive coverage, federal employees (ie, Federal Employees Health Benefits Program), Medicare recipients, and Medicaid insureds. Next, we totaled healthcare utilizations in the following 3 categories for each subpopulation: encounters with physicians and nonphysicians, admissions to hospitals, and days in the hospitals. In the Table, we show the membership totals of each of the 5 subpopulations.
Trends in the Number of Insureds, by Subpopulation, for the US Health Insurance Industry, 2006-2014
A quick review of the Table and Figure 2
(which corresponds to the Table) shows that the growth in Medicaid membership is as expected. For group health insurance, the trend clearly shows that the membership is on a decline. We also see an increase in individual health insurance, especially from 2013 to 2014, when the state insurance exchanges began. Both Medicare and Medicaid members tripled since 2005, while group insurance members declined by 35%. Figure 2 overlays trends in the number of insureds in our 5 subpopulations, as indicated in the Table.
Trends in Healthcare Utilization, by Subpopulation of Insureds, for the US Health Insurance Industry: Encounters With Providers
In line with Figure 2, Figure 3
(A and B) represents the changes in the usage of the medical delivery system as measured by the number of encounters with physicians and nonphysicians. The trend is similar to the growth and decline in members. The substantial increase in use by the Medicaid population began before the ACA in 2010 and the 2008 financial crisis. We see a decline in encounters for group insurance plans and an indication of growth in individual coverage since 2013, the beginning of the exchanges. Independently of the ACA, the increase in the aging population and the retirement of baby boomers contribute to the growth in encounters for the Medicare population.
Trends in Healthcare Utilization, by Subpopulation of Insureds, for the US Health Insurance Industry: Inpatient Admissions and Duration of Hospitalization
(A and B) shows that utilization of hospital admissions and duration of hospitalization by the subpopulation covered by comprehensive group plans is in decline. The Medicaid population has shown a dramatic growth in the use of these utilization factors. The results found by Brandt et al4
in “Methodological Effects on the Measurement of Repeat Hospitalizations” are supported by the trend in admissions for the Medicare population.
Generally, Figures 2-4 show that discernible trends in utilization were in place before the ACA. Moreover, the trends are substantially different among the 5 covered subpopulations that we identified. Particularly striking are 2 trends: 1) a decline in utilization, as well as in the number of covered members, for individuals covered by employer group health insurance; and 2) a strong increase in utilization for individuals covered by Medicaid and Medicare—especially Medicaid.
In addition, the following trends are less striking, but still notable: 1) growth in individual health insurance from 2013 to 2014 with the creation of the exchanges; and 2) federal employee coverage and individual comprehensive coverage also show increases in utilization and number of individuals covered, but these trends are modest in comparison with the dramatic changes for employer group insurance coverage and Medicaid and Medicare.
In seeking possible explanations for the trends we found, we must look beyond the ACA, as these trends were in evidence before the ACA was enacted. The ACA may or may not have enhanced the trends, but other factors were also in play, one surely being the Great Recession during the late 2000s, which reduced employment and cut off former employees from access to group insurance. In addition, the Great Recession led to the failure of businesses and consequent cancellation of group coverage plans; however, the recovery of employment in 2010 did not lead to a reversal of the downward trend in group coverage or utilization.
It can be hypothesized that the mandates and requirements of the ACA, as well as its offer of exchanges as alternatives to employer insurance, may have encouraged many employers to cancel their group insurance plans after 2010. It would be speculative, and perhaps pejorative, to suggest that the reduction of employer involvement in healthcare was an intended consequence of ACA. Nonetheless, employer group coverage has declined, for whatever reasons—intended or not—in spite of an increase in employment since 2010. For insureds covered by employer group insurance, the decline in utilization, such as encounters with providers, admission to hospitals, and days in the hospital, may also be attributed to the move from richer health insurance plans to more deductibles, coinsurance, and greater out-of-pocket expenses. Being in a bronze plan, as opposed to a “Cadillac” (rich) plan, can lower the use of the healthcare system, as well. Further details are provided in eAppendix 1.
In addition, the steady aging of the population, as the large baby boomer generation retires from the work force and joins the Medicare ranks, surely adds to the growth of utilization in this subpopulation. Steady advances in medical science and technology may also play a role in utilization growth, as illnesses that were formerly neglected or treated poorly become amenable to effective treatment.
Our discussion of employers’ group health insurance and eAppendix 1 have benefited from discussions with Hannah Clinger, a vice president with Willis and Willis.