Analysis of High Deductible Health Plans

Spending

Increased cost sharing by employees will reduce health care spending:

• Based on strong evidence from the literature, we expect that increased cost sharing will result in some reductions in health spending. Read more below
• Multiple studies confirm that individuals use less health care when faced with health plans requiring higher cost sharing, such as HDHPs. Read more below
• Health savings accounts in conjunction with HDHPs may blunt the decreases in health spending associated with higher cost sharing in health plans. Read more below
• Lower spending observed in HDHPs may be the result of favorable selection, that is, may attract a higher proportion of healthier enrollees. Read more below

Based on strong evidence from the literature, we expect that increased cost sharing will result in reductions in health spending.

The goal of increasing the number of employees in high deductible health plans (HDHPs) is to make enrollees more cost conscious about health care because they are paying more out of pocket than in traditional health plans. To meet the federal standards for an HDHP in 2009, the deductible must be at least $1,150 for an individual and $2,300 for a family. In theory, such consumers will demand higher quality services at a lower cost and use less unnecessary care. As a result, overall health care spending would be expected to decrease. An increased emphasis on consumer choices is also expected to encourage cost and quality competition among health care providers, resulting in lower prices for services.

The gold standard study of consumer cost sharing is the RAND Health Insurance Experiment (HIE), a randomized, controlled study conducted in the 1970s, which showed that increased consumer financial responsibility reduced both health care use and spending (Keeler et al., 1982). Health care spending among individuals in plans with the highest deductible was approximately 66 percent of spending among individuals in plans with no cost sharing requirements (Manning and Newhouse, 1987). Notably, decreased spending was due to reductions in use rather than a result of consumers shopping for the lowest cost care (Keeler, 1992). Individuals in plans with the highest level of cost sharing were 23 percent less likely to be hospitalized (Newhouse, 2004). The RAND HIE also found that individuals with the highest level of cost sharing reduced their use of services across the board by approximately 30 percent (Newhouse and Insurance Experiment Group, 1993).

Findings from the RAND HIE may not directly generalize to changes in spending under high deductible health plans. The HIE guaranteed that participants would not be financially disadvantaged by participating in the experiment. For example, if participants were randomly assigned to a health plan that had a lower value than their current coverage, they would receive monthly payments equal to the difference so that they would not be worse off in the new plan. It is unclear whether this arrangement adequately reflects the tax advantaged savings accounts of today's HDHPs.

Today, HDHPs are often combined with some form of savings option to allow people to set aside pretax dollars to be used to meet out-of-pocket health care expenses. The savings options take several forms, including tax free individual health savings accounts (HSAs) and employer sponsored HRAs. HDHPs in combination with some form of tax free savings account are known as consumer directed health plans (CDHPs).

Most research on the relationship between increased use of CDHPs and health care spending shows a reduction in spending. Buntin and colleagues (2006) reviewed the early evidence regarding the effect of consumer directed health care on health care spending, use, and quality and found that use decreased and spending increased at a slower pace under CDHPs than under other plans.

In 2001, Humana introduced several types of CDHPs among other traditional preferred provider organization (PPO) and health maintenance organization (HMO) health plans to its employees. During the first year, participation in CDHPs was 6 percent. CDHPs tended to be selected by healthier enrollees; costs per member per month for CDHP enrollees were half of the average for all employees in the year before the option was offered. Compared with prior experience estimates, CDHP enrollees' spending was 25—35 percent lower than expected, suggesting that these enrollees were more cost conscious in their health care decisions (Bertko, 2004; Hahn, 2005).

A recent Milliman study examining 30,000 CDHP enrollees with six employer programs found that, after adjusting for member risk profile and benefit factors, CDHP plans showed cost savings of approximately 1.5 percent compared with non-CDHP plans (Burke and Pipich, 2008). In their comparative study of spending and use among CDHP enrollees, Parente, Feldman, and Christianson (2004b) found significant increases in hospital costs and admission rates among CDHP enrollees compared with PPO and HMO enrollees. In a follow-up study, the authors found that CDHP enrollees had higher total expenditures, physician expenditures, and hospital expenditures than did HMO enrollees in 2001, 2002, and 2003. Conversely, CDHP enrollees had lower prescription drug spending (Feldman, Parente, and Christianson, 2007). Some of the variation in results of these studies may stem from the heterogeneous benefit designs of both the HDHPs and the conventional plans studied.

To estimate the effects on health care spending when people switch to HDHPs, a number of studies have used modeling and microsimulation techniques. Keeler et al. (1996) used HIE data with the RAND Health Expenditure Simulation Model to simulate changes in health care spending associated with HDHPs coupled with medical savings accounts (MSAs), an early form of tax advantaged savings account that has been mostly superseded by the HSA. The authors estimated that if all non-elderly insured individuals were to enroll in a health plan with an MSA, health care spending would decline by 0 to 13 percent. They also determined that many insured individuals would retain their current coverage, resulting in an overall change in health care spending of -1 to +2 percent. Baicker, Dow, and Wolfson (2006) used price sensitivity estimates (estimates of how demand for or spending on health services changes with price) at different points on a spending curve for a PPO enrollee who is in average health. After simulating the behavioral changes in response to different coinsurance rates, the authors found that switching from a typical PPO to a typical HSA/HDHP would, on average, reduce individual health spending by 5 percent in the short run. The simulation projected that if 28 million randomly selected people were to switch from a PPO to an HSA/HDHP by 2010, health care spending could decrease by 0.8 percent, or $6 billion per year (in 2006 dollars).

Nichols, Moon, and Wall (1996) modeled how MSAs coupled with catastrophic health plans (another term for an HDHP) would affect national health spending, consumer financial risk, health insurance premiums, and coverage. Using baseline health spending survey data weighted to reflect the nation as a whole, they estimated that if all workers switched to an MSA plus a catastrophic health plan, national health care spending could be reduced by 15 percent. However, after accounting for other confounding factors-such as the presence of public insurance, selection patterns, and the influence of the MSA on consumer decisionmaking-they found that reduction in spending would likely be in the range of 4 to 6 percent. In a cost-benefit analysis of MSA/catastrophic health plans, Zabinski and colleagues (1999) found that increased enrollment in medical savings accounts coupled with catastrophic health plans would lower administrative burden by approximately $100 per family and would lower family health care spending by $295 per family (or $14.7 billion).

HDHPs are relatively new and have low market penetration, so analyses of the relationship between HDHPs and spending are representative of a rather new trend in the health insurance market. As of 2007, approximately 2 percent of insured individuals had a CDHP and 11 percent had an HDHP (Fronstin and Collins, 2008). This market, however, has grown rapidly, increasing threefold between 2005 and 2007, from 1.0 million to 4.5 million (Center for Policy Research, 2007; America's Health Insurance Plans' survey of U.S. health insurance carriers did not include insurance plans coupled with HRAs, which would likely underestimate the enrollment rate in such plans).

Multiple studies confirm that individuals use less health care when faced with health plans requiring higher cost sharing, such as HDHPs.

Recent observational research examining the relationship between HDHPs and health care use has generally shown a decrease in use across many types of health care. Wharam and colleagues (2007) found a 10 percent reduction in total emergency department visits among HDHP enrollees, which they attributed primarily to a statistically significant 25 percent reduction in low severity repeat visits. The study also reported that a disproportionate share of the reduction in high severity first time visits occurred in the two lowest income groups (a 25 percent reduction) compared with the two highest income groups, in which there was a 1.3 percent reduction.

Gerfin and Schellhorn (2006) evaluated the effect of varying deductibles within the Swiss health care system on the probability of visiting a doctor. They found that higher deductibles were associated with a lower probability of going to the doctor, with approximately one-third of the reduction caused by increased price consciousness among patients and the remaining two-thirds attributable to the fact that healthier patients were more likely to choose high deductibles.

Health savings accounts in conjunction with HDHPs may blunt the decreases in health spending associated with cost sharing in health plans.

HDHPs may be coupled with health related spending/savings accounts (such as individual HSAs), employer sponsored HRAs, medical savings accounts, or flexible spending accounts (FSAs) (which allow the account to be used for medical as well as other specified expenses, such as dependent care). These accounts may blunt expected reductions in spending because they reduce incentives for consumers to constrain spending. Buntin and colleagues (2006) found evidence suggesting that the presence of a personal savings account (such as an HSA) attached to an HDHP could offset the overall spending reduction by roughly one-half, resulting in a net reduction in use of 2 to 7 percent.

Other factors may also offset the reduction in spending associated with plans involving high levels of cost sharing. Some insured individuals with substantial health care needs are already subject to significant cost sharing requirements; this group might actually see a reduction in their financial liability under a consumer directed health plan, which would reduce their price sensitivity. In addition, many HDHPs do not require cost sharing for preventive services, limiting decreased spending for those services.

Lower spending observed in HDHPs may be the result of favorable selection, that is, may attract a higher proportion of healthier enrollees.

A major limitation of studies comparing health care spending between HDHPs and more traditional plans is that they cannot completely control for the possibility that individuals who enroll in HDHPs are healthier overall than those who choose traditional plans. As a result, some portion of lower health care spending may occur because HDHP enrollees are healthier on average than enrollees in traditional plans.

HDHPs can vary significantly in the relative contribution to premiums of employers and employees, and these variations may affect overall health spending. If HDHPs are offered as a choice among several different types of health plans, they may tend to attract younger and healthier individuals, who have low health care needs, or high income employees, who can afford greater out-of-pocket costs in exchange for lower premiums. If healthier employees shift to HDHPs, comprehensive plans may increase premiums to account for the higher expected costs among the remaining enrollees. Such adverse selection might be ameliorated or eliminated if the employer subsidizes the costs of the comprehensive plan or if the employer offers only an HDHP.

In a case study of a company that decided to offer CDHPs in addition to more comprehensive plans, Tollen, Ross, and Poor (2004) found that new CDHP enrollees were substantially healthier before enrolling in the plan than were enrollees in other plans. CDHP enrollees used approximately 50 percent fewer health care services in the year before the study than did the company's employee average. The authors concluded that although healthier employees were more likely to enroll in the HDHP, this did not explain all of the decreased use. Humana found that enrollees in CDHPs spent 50 percent less than the average spending per member per month of all enrollees. CDHP enrollees also tended to have a higher income (Bertko, 2004). Barry et al. (2008), in analyzing Alcoa's decision to introduce a CDHP as a health plan choice to its employees, found that families that had a member with diabetes, asthma, or heart disease were less likely to enroll in the CDHP than families without a member with a chronic condition.

Greene et al. (2006), in a similar study investigating the characteristics of employees in a large firm who voluntarily enrolled in a CDHP, also found that employees who chose to enroll in a CDHP were younger, wealthier, and healthier than were workers who remained in more traditional plans. In their review of the evidence base on CDHPs, Buntin and colleagues (2006) found that few studies observed a difference in CDHP enrollees across age groups; however, those who enrolled in a CDHP were slightly more likely to have higher household incomes and to be in better health than those in other plans. They concluded that CDHPs are somewhat more likely to attract a healthier population. On the other hand, findings from a study of University of Minnesota employees' enrollment decisions after the university began offering CDHPs showed no significant difference in health between employees who chose to enroll in a CDHP and those who chose standard plans. However, the study did find that higher income employees and employees who preferred a larger choice of providers opted for CDHPs (Parente, Feldman, and Christianson, 2004a).

Back to top

Consumer Financial Risk

The effect of HDHPs on consumer financial risk will depend on the consumer's baseline health and level of health care spending:

• The effect of HDHPs on consumer financial risk will depend largely on the individual's level of health care spending. Read more below
• Several analyses suggest that consumers with few health care needs will see savings if they switch from a standard plan to an HDHP, whereas those with chronic diseases and moderate health care needs will likely face higher out-of-pocket costs for health care. Read more below
• Low income individuals may be disproportionately burdened by the cost sharing associated with HDHPs. Read more below

The effect of HDHPs on consumer financial risk will depend largely on the individual's level of health care spending.

Although HDHPs offer protection against catastrophic risk with lower premiums, consumers may be at higher risk of financial burden from the high upfront out-of-pocket payments than enrollees in more-comprehensive plans. All enrollees will save on premiums; however, individual consumer financial risk will depend on the level of health service use.

Studies have shown that switching to an HDHP plan that qualifies for a health savings account (HSA) could offer health care savings for some consumers, especially those with few health care needs, whereas low income and less healthy individuals would be most at risk for experiencing financial burden under HDHPs. Results from a simulation of out-of-pocket payments for premiums and medical care showed that young, healthy individuals benefit the most from CDHPs, but that moderately sick and low income individuals are most likely to experience significant financial burden from high out-of-pocket payments (McNeill, 2004).

Nichols, Moon, and Wall (1996) analyzed the effect of medical savings accounts (MSAs)—an earlier form of HSA—coupled with catastrophic health plans (HDHPs) on consumer financial risk and premiums. They found that switching to an MSA catastrophic health plan would save individuals, on average, $340 in out-of-pocket premium payments during the first year. Moreover, during the first year, "winners" would save approximately $402 per person; those who would be made worse off would spend an additional $397. The authors concluded that savings would be greater the less an individual spends on health care, suggesting that people with greater health care needs were likely to spend the most under an MSA catastrophic health plan. MSAs, however, were a temporary phenomenon and are no longer prominent in the HSA market. Thus, evaluations of MSAs may not be directly applicable to today's HSAs.

HDHPs may shift care away from hospitals and specialists and toward primary care physicians. For example, the RAND Health Insurance Experiment (HIE) found that individuals in the plan with the highest level of cost sharing were 23 percent less likely to be hospitalized. This reduction may be, in part, because physician visits and hospital care are complements: Some hospital care results from an office visit with a physician; if office visits are reduced, hospitalizations may decrease as well (Newhouse, 2004). A study of Humana's experience implementing CDHPs among other, more traditional plans found that hospital inpatient admissions declined, hospital outpatient visits were unchanged, and office visits increased (Bertko, 2004).

On the other hand, in their study of spending and utilization among CDHP enrollees compared with preferred provider organization (PPO) and point-of-service (POS) enrollees, Parente, Feldman, and Christianson (2004) found significant increases in hospital costs and admission rates among those enrolled in CDHPs; the researchers hypothesized that perhaps that CDHP enrollees waited until they were sicker to seek care and thus were more likely to require hospitalization.

Several analyses suggest that consumers with few health care needs will see savings if they switch from a standard plan to an HDHP, whereas those with chronic diseases and moderate health care needs will likely face higher out-of-pocket costs for health care.

Based on strong evidence from the literature, we would expect that individuals with low health care costs and those with very high health care costs could be made better off financially under current HDHP designs. Those with the highest health care costs (exceeding the HDHP's out-of-pocket maximum) may be at the same financial risk or slightly lower risk than they would be under traditional plans. In contrast, those with moderate spending, including many of the chronically ill, would be worse off.

Increased cost sharing under HDHPs will decrease the financial burden on younger, healthier employees because they have the lowest out-of-pocket spending. Under the current health care system, younger, healthier individuals enrolled in a comprehensive plan are likely purchasing more health insurance than they need. Thus, enrolling in an HDHP would lower their out-of-pocket health insurance purchasing costs, which are deducted from their payroll. Individuals with very extensive health care needs and high health spending may also realize health care savings under an HDHP because HDHPs generally have lower out-of-pocket maximum liabilities than traditional plans. On the other hand, individuals with medium to high levels of health care spending are likely to be worse off financially under an HDHP.

Baicker, Dow, and Wolfson (2006) simulated the effect of HSAs on spending and found that switching from a traditional PPO to an HSA would, on average, reduce individual health spending by 5 percent. However, the change in personal health spending varies according to individual health care needs. The authors show that individuals at the low and high ends of the spending spectrum would be either the same or better off in an HSA, whereas those in the middle range of spending (those most likely to have a chronic condition) could have increased expenses, as high as $545 per year (Baicker, Dow, and Wolfson, 2006, 2007).

An analysis of expected out-of-pocket payments in proportion to income under different deductible scenarios demonstrated that an individual's probability of experiencing financial hardship varies according to health. The study predicted that, under a $2,500 deductible plan, 13 percent of enrollees would have out-of-pocket payments greater than 10 percent of their income. However, 22 percent of individuals with a chronic condition, 53 percent of those with poor health, and 66 percent of individuals who were hospitalized would have out-of-pocket costs exceeding 10 percent of income (Trude, 2003).

The Employee Benefits Research Institute (EBRI)/Commonwealth Fund's Consumerism in Health Care Survey showed that 42 percent of respondents in HDHPs and 31 percent of those in CDHPs spent 5 percent or more of their income on deductibles, copayments, and premiums compared with 12 percent of those in more comprehensive health plans (Fronstin and Collins, 2008). The Commonwealth Fund's Biennial Health Insurance Survey showed that 54 percent of individuals in health plans with a deductible greater than $1,000 reported difficulty paying medical bills compared with 39 percent of individuals with deductibles under $500 and 24 percent of individuals in plans with no deductibles (Davis, Doty, and Ho, 2005). Although Wharam and colleagues (2007) found a reduction in emergency department use among CDHP enrollees compared with traditional plan enrollees, total out-of-pocket spending by CDHP enrollees increased substantially.

However, Keeler and colleagues (1996), who simulated the effect of HDHPs coupled with MSAs on health care costs, found that enrollees in HDHPs had lower per capita expenditures compared with enrollees in HMOs or fee-for-service plans. Moreover, Remler and Glied (2006) argued that the health insurance market, excluding HDHPs and CDHPs, already has substantial cost sharing and cost control mechanisms (e.g., utilization review and gatekeeping), suggesting that, especially for individuals most in need of medical care, the level of cost sharing either would not change or might even be reduced with an expansion of HDHPs.

An American Academy of Actuaries (2004) report on CDHPs demonstrated that, if health care use is held constant, those with few health care needs will save $360 in a CDHP compared with a PPO, whereas moderate and high health care users will spend an additional $615 and $650, respectively, in a CDHP. Further, the report estimated that if a consumer reduces health care use modestly by decreasing non-preventive office visits and associated labs (a 3 percent reduction in physician costs), he or she will save about $10 in out-of-pocket costs (2 percent). A more significant reduction in health care use by changing care from inpatient to outpatient settings, using more efficient providers, and decreasing hospital stays (a 7 percent reduction in physician costs and a 2 percent reduction in hospital costs) will decrease the consumer's out-of-pocket costs by $20 (4 percent).

If adverse selection leads a disproportionately sicker population to remain in comprehensive health plans, enrollees in those plans could see premiums increase. As healthier individuals migrate to HDHPs, then plans with more comprehensive benefits will likely raise premiums for their remaining beneficiaries to cover the increased expected cost of the pool of enrollees. Premiums for comprehensive plans could spiral up. Age profiles of those enrolled in HDHPs and those in more traditional plans are similar, but HDHP enrollees tend to be slightly healthier.

Low income individuals may be disproportionately burdened by the cost sharing associated with HDHPs.

Low income individuals and those with health problems would be placed at a disadvantage under a HDHP because a larger portion of their income would be spent on health care compared with higher income or healthier individuals. The Commonwealth Fund's Biennial Health Insurance Survey showed that individuals in plans with high deductibles who had low incomes or health problems were more likely to experience medical debt or difficulty paying medical bills than were similar individuals in plans with low or no deductible (Davis, Doty, and Ho, 2005).

The existence of HSAs could offset some of the increased consumer financial risk that results from increased cost sharing. However, if an HSA is attached to the HDHP, tax advantages associated with HSAs primarily benefit those who pay a higher marginal income tax rate (the tax rate that applies to the last dollar of income, i.e., the change in tax rates as income increases), thus benefiting higher income individuals the most. In addition, one study (Minicozzi, 2006) has shown that, overall, such accounts have a relatively short tenure, and higher income individuals are more likely to contribute to them for a longer period of time than are lower income individuals. Minicozzi showed that the majority of individuals with an MSA used their account as a savings mechanism rather than as a spending account; approximately 65 percent of individuals with an MSA spent less than half of the amount that they contributed. However, the study also found that approximately one-third of MSAs lasted for at least three years, whereas nearly half lasted for less than one year. Higher income individuals were more likely to have an MSA that lasted for more than one year (Minicozzi, 2006).

Back to top

Waste

The effect of increased cost sharing on waste is uncertain:

• There is no direct evidence on the relationship between HDHPs and waste in the health care system. Read more below
• In general, a higher level of cost sharing leads to reduced use of medical services; it can be difficult to determine whether these services constitute necessary or unnecessary (i.e., wasteful) care. Read more below

There is no direct evidence on the relationship between HDHPs and waste in the health care system.

No literature directly examines the effect of plans with high levels of cost sharing, such as HDHPs, on waste in the health care system. Most of the research focuses on the effects on health care use in general, without specifically identifying unnecessary care.

In general, a higher level of cost sharing leads to reduced use of medical services; it can be difficult to determine whether these services constitute necessary or unnecessary (i.e., wasteful) care.

HDHPs are intended to make enrollees more discerning consumers of health care services, that is, consumers who will demand higher quality services at a lower cost and use less unnecessary care. More cost and quality competition could also encourage providers to become more efficient. However, the reductions in use that result from high cost sharing may include necessary care as well as wasteful care, and it can be difficult to distinguish between them.

Multiple studies confirm that high cost sharing reduces the use of all types of health care. The RAND Health Insurance Experiment (HIE) found that individuals with the highest level of cost sharing reduced their use of services across the board by approximately 30 percent — especially low income and less healthy adults and children (Newhouse and Insurance Experiment Group, 1993). Keeler and colleagues (1996), in their study on the health spending effect of medical savings accounts (MSAs), an earlier form of tax advantaged savings account, found evidence that HDHPs coupled with MSAs could reduce excess medical care use among the insured, depending on the magnitude of the deductibility. In their study of the relationship between HDHPs and emergency department use, Wharam and colleagues (2007) found a 10 percent reduction in total emergency department visits, which they ascribed primarily to a statistically significant 25 percent reduction in low severity repeat visits. The study also reported that a disproportionate share of the reduction in high severity first time visits occurred in the two lowest income groups (a 25 percent reduction) compared with the two highest income groups, in which there was a 1.3 percent reduction.

Because HDHPs are relatively new to the insurance market, we do not yet fully understand how they will change consumer health seeking behavior. For example, we do not yet know whether consumers will use more preventive and primary care in order to avoid high costs in the longer term, reduce use of non-essential medical care, or will delay or avoid seeking care altogether because of high out-of-pocket payments. We also do not know whether the health system will respond to an increased emphasis on consumerism with improvements in efficiency.

Back to top

Reliability

Cost sharing reduces both necessary and unnecessary care, so its effects on reliability of care are unclear:

• There is no direct evidence from the literature on whether HDHPs will improve the quality of health care delivered to enrollees. Read more below
• Higher cost sharing reduces health care use of both necessary and unnecessary care; therefore, effects on the reliability of care may be mixed. Read more below
• Multiple studies show that cost sharing leads to lower rates of preventive care and health screening exams; however, some features of HDHPs may encourage preventive care. Read more below

There is no direct evidence from the literature on whether HDHPs will improve the quality of health care delivered to enrollees.

The main mechanism by which HDHPs may affect reliability is through health delivery system performance. If such performance improves, this could attract discerning consumers who are shopping for high quality services at competitive prices. However, given the presently low market penetration of HDHPs, we would not expect a generalized higher response to this type of health plan.

Higher cost sharing reduces health care use of both necessary and unnecessary care; therefore, its effects on the reliability of care may be mixed.

Increased employee cost sharing is intended to encourage enrollees to use available cost and quality information to make more cost conscious health care decisions, seek out higher quality care, and reduce the use of unnecessary care. However, increased cost sharing may discourage individuals from seeking essential as well as non-essential medical care and thus may decrease the reliability of care.

Lower income individuals, faced with a large deductible, may be more likely to defer or avoid needed care because of the financial barriers to access. Individuals in poor health may also face challenges in getting needed care because of the volume of health care services they require and the greater financial burden these services may pose under a high deductible plan. Existing racial and socioeconomic disparities in access to care could be exacerbated by increased cost sharing requirements if such subgroups were to enroll in HDHPs.

Research has shown that cost sharing reduces the likelihood of patients receiving either appropriate or inappropriate care, especially among low income and less healthy adults and children. The RAND Health Insurance Experiment (HIE) found that higher cost sharing led to reductions in both essential and non-essential medical care use. Cost sharing reduced the probability of using preventive services, such as cancer screening, as well as appropriate hospital admissions. Cost sharing reduced the use of Pap smears from 65 percent to 52 percent and reduced the use of appropriate hospital admissions by 22 percent (Lurie et al., 1987; Siu et al., 1986; Rosenthal, Hsuan, and Milstein, 2005). The RAND HIE showed that the probability of receiving effective medical care among low income children and adults in plans with cost sharing mechanisms was nearly half that among those in plans without cost sharing (Lohr et al., 1986). Results from a more recent study of mammography use among women age 65 to 69 enrolled in Medicare managed care plans that did and did not require cost sharing for biennial breast cancer screening showed an 8.3 percentage point lower screening rate among women in plans with cost sharing requirements. Lower income and less educated women were disproportionately affected by the cost sharing requirements (Trivedi, Rokowski, and Ayanian, 2008).

A number of studies confirm that cost sharing within CDHPs reduces use of medical care. Parente, Feldman, and Christianson (2004) found that CDHP enrollees initially had fewer physician visits and lower use of prescription drugs. However, hospital admissions doubled three years later, indicating that underuse of essential medical care to prevent, diagnose, or control conditions at an early stage may have resulted in increased use of more intensive and more expensive care later. The authors noted that the increased hospital admissions may have been caused by financial disincentives for employees initially in more traditional HMOs and PPOs to use a particular medical center, encouraging them to switch into the CDHP later and thus use their preferred, more expensive care setting.

Research has shown that high levels of cost sharing under HDHPs may cause enrollees to delay, skip, or avoid seeking needed medical care in order to keep out-of-pocket spending low. Results from the EBRI/Commonwealth Fund's Consumerism in Health Care Survey show that approximately one-third of individuals with CDHPs or HDHPs reported delaying or avoiding care, whereas only 16 percent of those in more comprehensive health plans reported the same. Skipping, delaying, or avoiding care was especially apparent among CDHP and HDHP participants with annual incomes below $50,000, those with fair or poor health, and those with at least one chronic condition (Fronstin and Collins, 2008). Findings from a survey of HDHP enrollees showed that, due to cost, respondents were more likely not to fill a prescription, more likely to forgo preventive services, and more likely to avoid seeing a physician than were those in other types of health plans (Lee and Zapert, 2005).

Gerfin and Schellhorn (2006) evaluated the effect of varying deductibles within the Swiss health care system on the probability of visiting a doctor. They found that higher deductibles were associated with a lower probability of going to the doctor, with approximately one-third of the reduction caused by increased price consciousness among patients and the remaining two-thirds attributable to healthier patients being more likely to select high deductibles.

Wharam and colleagues (2007) found a 10 percent reduction in total emergency department visits, which they attributed primarily to a 25 percent statistically significant reduction in low severity repeat emergency department visits. The study also reported that a disproportionate share of the reduction in high severity first time visits occurred in the two lowest income groups (a 25 percent reduction) compared with the two highest income groups, in which there was a 1.3 percent reduction.

Multiple studies show that cost sharing leads to lower rates of preventive care and health screening exams; however, some features of HDHPs may encourage preventive care.

The design features of HDHPs may include incentives to encourage use of preventive care and disease management programs, thus avoiding the cost disincentive to forgo care. The Internal Revenue Code allows HDHPs to provide preventive care benefits without a deductible or with a deductible below the minimum annual deductible. Such preventive benefits include routine health care visits, immunizations, certain health promotion programs, as well as screenings for cancer, chronic care, and mental health (U.S. Department of the Treasury, 2004).

Because most HDHPs cover 100 percent of preventive care before the deductible is met (Center for Policy Research, 2007), a number of studies examine the effect of HDHPs on the use of preventive services. Rowe et al. (2008), in a study of the use of preventive services, cancer screening, and use of diabetes preventive care, showed that enrollees in CDHPs had the same or higher use of such preventive services than did those enrolled in a PPO. Fronstin and Collins (2008) found that, although HDHP enrollees were more likely to skimp on needed care than were enrollees in more comprehensive plans, the rates of preventive tests and services did not differ with enrollees in more comprehensive plans.

In contrast, Busch and colleagues (2006) compared use of preventive care between employees who were affected by an employer's decisions to remove cost sharing requirements (e.g., copayments and deductibles) for most preventive care services and employees who were not affected by the policy changes. They found that removing cost sharing requirements for preventive care did not affect use. For example, 37 percent of women who were subject to the policy change had a cervical screening compared with 38 percent of women who did not have a change in their health plan's cost sharing requirements. However, the authors warn that the lack of relationship may be due to the initially low cost sharing requirements for preventive care before the policy change.

Early results from a study of four large employers that implemented a full replacement of all health plans with CDHPs showed that the use of preventive services stayed the same or declined slightly (few of the declines in preventive services use were statistically significant). These preventive services were covered at 100 percent without any cost sharing at all of the firms (Parente et al., 2008). A McKinsey & Company (Agrawl et al., 2005) survey of employees whose employers replaced all health plans with a HRA/CDHP found that CDHP enrollees were 25 percent more likely to report that they engaged in healthy behaviors than members of traditional plans, 20 percent more likely to engage in wellness programs, more than 30 percent more likely to take advantage of preventive services (e.g., annual checkups), and 20 percent more likely to manage chronic conditions carefully. One limitation of this study is that it is based on survey data as opposed to actuarial data; therefore, it may be subject to faulty recall on the part of respondents.

A number of insurance industry studies also suggest that enrollees in HDHPs are more likely to seek preventive care and to engage in healthy behaviors. Although informative, these studies do not typically take into account the favorable selection of HDHPs by healthier individuals. An Aetna (2004) study of individuals before and after they enrolled in the company's HRA plan showed that, over a four year period, enrollees increased their use of preventive health care, and enrollees with chronic conditions maintained or improved their use of care. The use of preventive health care increased by 23 percent among adults enrolled in the HRA compared with an 8 percent increase among adults who did not enroll. HRA enrollees with diabetes either maintained their level of certain diagnostic tests and screening or increased use of other tests by 4 to 6 percent, depending on the test. Results from a Blue Cross Blue Shield survey (Sullivan, 2006) showed that 20 percent of its HSA members reported using a wellness program compared with 8 percent of those who were not enrolled in a CDHP.

Back to top

Patient Experience

Evidence from high cost sharing health plans suggests that enrollees lack adequate information to make health care decisions that may affect patient experience:

• Multiple studies suggest that employers and enrollees in high cost sharing plans feel that that they lack sufficient information on cost and quality to make health care decisions. Read more below
• Several surveys suggest lower consumer satisfaction with CDHPs and HDHPs than with traditional health plans. Read more below
• Despite the dissatisfaction with available cost and quality information, reenrollment rates in HDHPs remain higher than in traditional health plans. Read more below

Multiple studies suggest that employers and enrollees in high cost sharing plans feel that that they lack sufficient information on cost and quality to make health care decisions.

A major complaint among enrollees in HDHPs is the lack of available information to help make informed decisions about their care. HDHPs are typically combined with some form of tax advantaged savings account in a consumer directed health plan (CDHP). According to results from a McKinsey & Company survey of CDHP consumers, a majority of CDHP enrollees were dissatisfied with the information available through their plan, and the majority of those who reported dissatisfaction cited the lack of information about physician costs as their primary frustration with their plan. Approximately 80 percent of CDHP consumers reported that they did not have sufficient physician cost information (Agrawl et al., 2005). Davis (2004, p. 1227), in her review of studies of CDHPs' effects on patient satisfaction, noted "some indication in the studies that few people find the CDHP easy to understand or the Internet-based tools easy to use." Based on interviews with key representatives from four firms that offered their employees consumer directed health plans among other plans, Lo Sasso and colleagues (2004) found multiple company representatives citing the importance of educating employees about CDHPs prior to enrollment.

Buntin and colleagues (2006) found that more insurers and employers are beginning to offer their consumers and employees information about health care and tools to help them make decisions. Despite such efforts, few CDHP enrollees feel that they have enough information to make appropriate decisions. Results from an examination of more than 40 large employers that offer at least one CDHP showed that most employers are not confident about the information that they do provide to their employees, especially quality and cost information about hospitals and physicians. Only 7 percent of employers rated the information that they provided about the cost of health providers as "excellent" or "good," and only 10 percent rated the information they provided about quality as "good." Employers were more positive, however, about the health management programs that they offered to their employees and about information on prescription drugs (McDevitt et al., 2007).

Several surveys suggest lower consumer satisfaction with CDHPs and HDHPs than with traditional health plans.

Individuals enrolled in HDHPs are generally less satisfied with their plans than are those with more comprehensive health insurance. The Employee Benefits Research Institute (EBRI)/Commonwealth Fund's Consumerism in Health Care Survey found that only 47 percent of CDHP enrollees and 35 percent of HDHP enrollees stated being very or extremely satisfied with their plans, compared with 64 percent of those with more comprehensive insurance. Only 45 and 34 percent of CDHP and HDHP enrollees, respectively, reported that they were very likely to stay with their plans compared with 64 percent of those with more comprehensive coverage. The increase in the proportion of the population who reported being extremely or very satisfied with their health plan was statistically significant for those with CDHPs between 2006 and 2007 (Fronstin and Collins, 2008). Christianson, Parente, and Feldman (2004) found similar results from their study of University of Minnesota employees enrolled in CDHPs. Fewer than half of CDHP enrollees were satisfied with their plans, and only 30 percent were willing to recommend a CDHP to others.

The Commonwealth Fund's Biennial Health Insurance Survey found that approximately 38 percent of individuals in health plans with deductibles of $1,000 or more are likely to experience problems accessing the health care system compared with 30 percent of those with deductibles under $500 and 21 percent among those with no deductible. Survey results also showed that individuals with low incomes or health problems were more likely to experience access problems than similar individuals with low or no deductibles (Davis, Doty, and Ho, 2005).

Despite the dissatisfaction with available cost and quality information, reenrollment rates in HDHPs remain higher than in traditional health plans.

Despite reported dissatisfaction with HDHPs, studies show that reenrollment, another measure of consumer satisfaction, is high among participants in HDHPs. Approximately 90 percent of enrollees reenroll in HDHP (Davis, 2004).

Back to top

Health

Increased cost sharing does not seem to affect overall population health, but it could have negative effects on the poor and the sick:

• While evidence suggests that the health of the overall population may not change with increased cost sharing, the health of individuals with low income and greater health care needs may decline. Read more below
• A number of studies suggest that use of health services decreases with high cost sharing, but the effects on health remain uncertain. Read more below
• The design features of an HDHP can be used to encourage the use of preventive health services or to mitigate the effects of cost sharing for low income enrollees, but no studies show that such features lead to changes in health. Read more below

While evidence suggests that the health of the overall population may not change with increased cost sharing, the health of individuals with low income and greater health care needs may decline.

Health outcomes for the general population have not been shown to be affected by increased cost sharing requirements (e.g., copayments and deductibles); however, lower income and less healthy individuals in plans with high cost sharing tend to experience poorer health outcomes than do similar individuals who are subject to low or no cost sharing.

We found no studies describing a direct relationship between the current generation of HDHPs and clinical health outcomes. However, literature on the effects of cost sharing on health care use can shed light on some of the expected effects on health outcomes of changing patterns of health care use.

The RAND Health Insurance Experiment (HIE) found that reduced use of both essential and non-essential care did not lead to poorer health outcomes among the general population. However, it did find that low income individuals with identifiable health problems using cost sharing plans were less likely to have their blood pressure controlled and their vision corrected than similar people in plans that had no cost sharing (Lohr et al., 1986; Brook et al., 1983; Rassel, 1995).

A number of studies suggest that health service use decreases with high cost sharing, but the effects on health remain uncertain.

Most studies on changes in health service use with HDHPs do not evaluate health outcomes or may not have an adequate study time frame to capture resulting changes in health. Generalizations from the existing evidence base are also limited because of the low market penetration of HDHPs, the homogeneous characteristics of current HDHP enrollees, and the limited sample population of most studies. Finally, it can be difficult to determine the effect of risk segmentation on the health effects of HDHP; for example, if individuals with healthier lifestyles disproportionately choose HDHPs, the effect of HDHPs on health may be difficult to discern.

High cost sharing for prescription drugs, whether in traditional plans or HDHPs, has been shown to have adverse effects on adherence to recommended care. For some chronic conditions, including congestive heart failure and diabetes, increased cost sharing is associated with lower adherence, more frequent discontinuation, and increased use of other health care services (Goldman, Joyce, and Zheng, 2007). Greene et al. (2008) compared prescription drug use among enrollees in HDHPs and those in traditional plans and found that prescription drug use for asymptomatic chronic diseases (e.g., high blood pressure and high cholesterol) declined among those enrolled in the HDHP. The health effect of these changes has not been evaluated.

The design features of an HDHP can be used to encourage the use of preventive health services or to mitigate the effects of cost sharing for low income enrollees, but there are no studies that show that such features lead to changes in health.

If HDHPs require less cost sharing for preventive care than traditional plans, they could encourage the use of such care and potentially have a positive effect on health. Many HDHPs exempt preventive care from the deductible and actually require less cost sharing than traditional plans. According to America's Health Insurance Plans' (AHIP) Center for Policy Research (2007), the majority of HSA/HDHP plans allow coverage of preventive health care before meeting the deductible.

The literature shows mixed evidence on the use of preventive care. Rowe and colleagues (2008) compared enrollees in CDHPs with those in preferred provider organizations (PPOs) on their use of general preventive services, preventive care for diabetes, and cancer screening. The authors matched participants on age, sex, family status, geographic area, and risk of health service use. They found that enrollees in CDHPs had the same or greater use of such preventive services as those enrolled in a PPO. Busch and colleagues (2006) compared preventive care use between employees who were affected by an employer's decisions to remove cost sharing requirements for most preventive care services and employees who were not affected by the policy changes. They found that removing cost sharing requirements for preventive care did not affect use. For example, 37 percent of women who were subject to the policy change had a cervical screening compared with 38 percent of women who did not have a change in their health plan's cost sharing requirements. However, the authors warn that the lack of relationship may be due to the initially low cost sharing requirements for preventive care before the policy change.

Early results from a study of four large employers that implemented a full replacement of all health plans with CDHPs found that the use of preventive services stayed the same or declined slightly (few of the declines preventive service use were statistically significant). These preventive services were covered at 100 percent without any cost sharing at the firms (Parente et al., 2008). A McKinsey & Company (Agrawl et al., 2005) survey of employees whose employers replaced all health plans with an HRA/CDHP found that CDHP enrollees were 25 percent more likely to report that they engaged in healthy behavior than were members of traditional plans, 20 percent more likely to engage in wellness programs, over 30 percent more likely to take advantage of preventive services (e.g., annual checkups), and 20 percent more likely to carefully manage chronic conditions. One limitation of this study is that it is based on survey data as opposed to actuarial data; therefore, it might be subject to faulty recall on the part of respondents.

Aetna's (2004) study of individuals before and after enrollment in an HRA plan showed that, over a four year period, enrollees increased their use of preventive health care, and enrollees with chronic conditions maintained or improved their use of care. The use of preventive health care increased by 23 percent among adults enrolled in the HRA compared with an 8 percent increase among adults who did not. HRA enrollees with diabetes either maintained their level of certain diagnostic tests and screening or increased use of other tests by 4 to 6 percent, depending on the test. Results from a Blue Cross Blue Shield survey showed that 20 percent of its HSA members reported using a wellness program, compared with 8 percent of those who were not enrolled in a CDHP (Sullivan, 2006). A limitation of these insurance industry studies is that they may not take into account the favorable selection of HDHPs by healthier individuals who might be more likely to seek preventive care and screening regardless of health plan type.

Back to top

Coverage

There is considerable uncertainty about how increased cost sharing will affect coverage:

• No studies directly examine how a policy to increase HDHP enrollment in the employer based insurance market would affect coverage rates. Read more below
• The findings from the literature are mixed and sensitive to modeling assumptions. Read more below

No studies directly examine how a policy to increase HDHP enrollment in the employer based insurance market would affect coverage rates.

Although use of HDHPs has been growing rapidly, there is no comprehensive, reliable set of data from which to accurately determine how these plans will fare in the future. We do not yet know how expanding use of HDHPs in the non-group or the group health insurance market will affect the provision of comprehensive health plans or the decision by employers to continue offering health insurance.

Proponents of HDHPs argue that the low premiums and savings mechanisms under such arrangements would encourage individuals to save for future health care expenses and allow low income and uninsured people to purchase coverage, thus reducing the number of uninsured. In addition, because HDHPs are less expensive than comprehensive policies, more small businesses would be able to offer insurance to their employees. Supporters also suggest that HSAs will improve the portability of health insurance, thus improving the continuity of coverage and reducing "job lock."

In contrast, critics of HDHPs argue that these plans would place low income individuals and/or those with chronic health conditions at a financial disadvantage and reduce the value of having health insurance. This is because HDHP enrollees must pay the full cost of care until the deductible is reached; when faced with a high deductible, low income individuals may, in effect, be uninsured because they cannot afford to pay for care. In addition, the tax preferences of HSAs benefit higher income individuals more because (1) those individuals are in a higher tax bracket and (2) they can contribute large amounts to their HSAs, giving them a better tax break. Moreover, the risk of adverse selection—the shifting of healthier individuals into lower cost plans (e.g., HDHPs), leaving individuals who are less healthy in higher cost, generous plans—raises additional concerns among opponents. If this shift occurs, premiums for those who remain in more traditional, comprehensive plans will increase, potentially destabilizing existing risk pools and resulting in premium spirals. This situation could lead to some individuals losing coverage. Opponents have also argued that encouraging the growth of HDHPs coupled with HSAs could crowd out comprehensive coverage, increasing the number of uninsured who previously were covered by employer sponsored insurance.

The findings from the literature are mixed and sensitive to modeling assumptions.

Studies of how HDHPs will affect health insurance coverage are sparse, and the studies that do exist provide mixed results. Glied and Remler (2005) predicted that HDHPs would increase coverage by 0.3 to less than 2 percent. The authors hypothesized that, because many in the non-group market are already enrolled in plans with high enough deductibles to qualify as HDHPs, increasing the attractiveness and availability of HSA/HDHPs will probably not persuade the currently uninsured to purchase non-group coverage. In addition, since more than half of the uninsured have no tax liability, they would not realize any gains from the tax advantages of HSAs. Potential savings from HSA qualified plans compared with more generous plans range from $0 to $117 per year, depending on an individual's tax bracket. Individuals in lower tax brackets receive the lowest cost savings from HSA qualified plans.

Feldman and colleagues (2005) simulated insurance participation rates in relationship to tax subsidies for HSA qualified plans. They predicted that President George W. Bush's refundable tax credit for HSAs would cause enrollment in individual HSAs to double and reduce the number of uninsured by 2.9 million. Contrary to opponents of HSA/HDHPs, they suggested that the tax credit would have a minimal effect on employer sponsored coverage, estimating that approximately 1 percent of employees would substitute employer sponsored coverage for an individual, non-group HSA plan. The authors suggested that, if employers continue to offer more comprehensive plans, participation in HSA related plans might not grow to represent a large portion of the market. The tax treatment of employer contributions to health insurance is likely to provide a disincentive for employers to offer less comprehensive, lower cost plans to their employees. The authors warn that because such plans are so new to the market, it is not possible to accurately determine how the popularity of these plans will grow in the future and how policy changes will influence insurance rates.

In an analysis of the effect of HSAs on insurance coverage and costs, Parente and colleagues (2005) showed that the uninsured are responsive to the price of insurance and that providing tax credits to individuals with an HSA will encourage participation in related plans. The authors show, however, that such participation is greater among higher income individuals.

Opponents of offering tax deductions or credits for the purchase of consumer directed health plans provided in the individual non-group market argue that increasing the attractiveness of such plans could theoretically result in an erosion of the employer provided health insurance market and a reduction in the number of insured individuals. As more workers, especially healthy or higher income workers, substitute employer provided coverage for non-group HSA/HDHPs, employers are likely to drop coverage or provide only low cost plans (i.e., HDHPs), because the pool of workers left in employer provided plans is likely to be less healthy and more costly to cover. If so, unhealthy and lower income workers would be made worse off. Gruber (2006) predicted that President Bush's proposal to increase incentives for people to purchase HSA/HDHPs would increase the number of people with such plans by 8.3 million, of whom 3.8 million were previously uninsured. At the same time, he projected that 8.9 million people would lose employer sponsored health insurance because their employers (typically small firms) would stop offering coverage. This would result in a net increase in the number of uninsured of 600,000.

Hoffman and Tolbert (2006) argued that high out-of-pocket costs associated with HDHPs would discourage lower income families from purchasing such plans. Moreover, the authors argued, most uninsured individuals do not have incomes high enough to benefit from the tax advantages of participating in such plans and their correlated HSAs. They concluded that HDHPs are not likely to substantially reduce the number of uninsured.

Park, Friedman, and Lee (2003) argued that encouraging participation in HSAs will encourage employers to substitute high deductible plans for more comprehensive health plans. As noted above, adverse selection and associated increases in premiums for more comprehensive health plans could reduce the number of employers offering such plans or reduce the number of employees enrolling in them. Either scenario will reduce the number of insured workers.

The few studies that have estimated the effect of HSA/HDHPs on crowding out comprehensive coverage show mixed results. In their analysis of medical savings accounts (MSAs) coupled with catastrophic health plans, Zabinski and colleagues (1999) showed that as enrollment in MSA/catastrophic health plans increases, premiums for such plans will decrease and premiums for more comprehensive coverage will increase.

In their case study of a company that decided to offer CDHPs in addition to more comprehensive plans, Tollen, Ross, and Poor (2004) found that new CDHP enrollees were substantially healthier before enrolling in the plan than were enrollees in other plans: CDHP enrollees used approximately 50 percent fewer health care services in the year before the study than did the company's employee average.

On the other hand, findings from a study of University of Minnesota employees' enrollment decisions after the university began offering CDHPs showed no significant difference in health between employees who chose to enroll in a CDHP compared with those who chose a more standard plan. However, the study did find that higher income employees and employees who preferred a larger choice of providers opted for CDHPs (Parente, Feldman, and Christianson, 2004).

Studies have generally shown that enrollees in HSA/HDHPs are of higher socioeconomic status. In their examination of enrollees in flexible spending accounts (FSAs), Hamilton and Marton (2008) showed that non-white individuals are less likely to participate in an FSA, make smaller contributions to an FSA, and switch to an HDHP. Some studies have shown that HSA/HDHPs attract higher income more than lower income individuals. Minicozzi (2006) showed that individuals who have an MSA are more likely to be middle aged and more than twice as likely to have an income greater than $100,000 per year. More than half (58 percent) of individuals enrolled in an HSA qualified plan have incomes greater than $50,000 per year, and 22 percent have incomes greater than $100,000 per year (Clemans-Cope, Blavin, and Kenney, 2006).

Capacity

HDHPs may alter the use of different types of health services, but no direct effect on capacity is likely:

• No empirical studies have examined the relationship between health system capacity and increased cost sharing by employees. Read more below

No empirical studies have examined the relationship between health system capacity and increased cost sharing by employees.

No studies directly examine the effect of HDHPs on capacity, but in the short term we would not expect a direct effect of increased cost sharing on the number of facilities, beds, or personnel. Over the long term, some changes in capacity may occur as a result of market adjustments, but other intervening factors make it difficult to attribute such adjustments to this one policy change.

Back to top

Operational Feasibility

Plans with high cost sharing are relatively easy to implement, though the successful implementation of these policies is hampered by a lack of sufficient information on health care resources:

• Employer based HDHPs are, administratively, relatively easy to implement, although challenges for both employers and employees remain. Read more below
• If employers and employees lack sufficient price and quality information on HDHPs to make health care decisions, they may face difficulties in effectively using HDHPs. Read more below
• Plans with high levels of cost sharing may pose additional challenges to providers in collection of fees. Read more below

Employer based HDHPs are, administratively, relatively easy to implement, although challenges for both employers and employees remain.

Evidence has shown that most employers that have introduced an HDHP found them feasible to implement and expect more employees to enroll in such plans in the future. Lo Sasso et al. (2004) conducted a case study of four companies that had begun offering consumer directed health plans (CDHPs) to their employees. In interviews with key company representatives, they found that most employers reported little administrative burden associated with CDHPs compared with other types of plans they offered. However, the authors warned that few employees within the companies interviewed had enrolled in the CDHPs and noted that administrative burden could increase if more employees enrolled.

In the same study, employers frequently cited the importance of educating employees about CDHPs before offering such plans (Lo Sasso et al., 2004). A study of 42 employers that offered at least one CDHP showed that approximately 90 percent of employers found communication about the need for and logistics of a CDHP to be their biggest challenge in implementing the plans (McDevitt et al., 2007).

Despite concerns about the capacity of individuals in HDHPs to make health care decisions, such plans have been growing in popularity. Enrollment in high deductible health plans increased threefold between 2005 and 2007, from 1.0 million to 4.5 million (Center for Policy Research, 2007; America's Health Insurance Plans' survey of U.S. health insurance carriers did not include insurance plans coupled with health reimbursement accounts (HRAs), which would likely underestimate the enrollment rate in such plans). Enrollment in health savings accounts (HSAs) or HRAs—tax preferred savings accounts typically coupled with HDHPs—is expected to reach 6.1 million by January 2008 (The Consumer Driven Health Care Institute, 2007).

If employers and employees lack sufficient price and quality information on HDHPs to make health care decisions, they may face difficulties in using HDHPs.

Research has shown that few health plans of any type provide appropriate information for consumers. Results from a study of 42 employers show that very few employers were satisfied with the information about cost and quality that was provided to their employees. Employers were more satisfied with information about prescription drugs and the health management programs that they offered (McDevitt et al., 2007). Only 7 percent of employers rated the information about cost as good or excellent, and 10 percent rated the information about quality as good.

Employee views of the quality of information available to them are consistent with the perspective of employers. For example, the Employee Benefits Research Institute (EBRI)/Commonwealth Fund's Consumerism in Health Care Survey found that fewer than one-third of adults in health plans of any type reported that their health plan provided cost and quality information about providers, and few health plan enrollees have trust in the information provided by their health insurance carrier (Fronstin and Collins, 2008). A McKinsey & Company survey of insured individuals, which included both CDHP members and members in other, more traditional types of plans, found that approximately 80 percent of CDHP consumers said they did not have sufficient information about physicians' charges (Agrawl et al., 2005). Rosenthal, Hsuan, and Milstein (2005) argued that one of the major gaps in information is the lack of comparison of appropriately detailed quality and cost measures across multiple providers and settings. Although America's Health Insurance Plans' recent survey of health insurance providers showed that the majority of insurers who offer HDHPs provide some quality and cost information, only half of the companies surveyed provided physician specific quality data. This survey did not assess the usability of the information by consumers (Center for Policy Research, 2007).

The success of CDHPs assumes that participants use health information to make well informed decisions about their health care. In their review of existing literature about CDHPs, Buntin and colleagues (2006) found evidence that CDHP members were more likely than consumers in more traditional plans to research information about their plan coverage, the cost and quality of care, and alternative care options. Many, however, continue to rely on their health care providers and social networks for assistance in making medical care decisions (Taylor, 2002). The Community Tracking Study (Trude and Conwell, 2004) reported concerns that employees were not well versed in making complicated health care decisions, regardless of the level and quality of information available. Some employers were especially concerned that employees with language barriers, lower educational attainment, and minimal access to the Internet would not be able to benefit from a CDHP. However, in a PricewaterhouseCoopers (2005) survey, 80 percent of chief executive officers felt that their employees could make wise decisions given necessary information. Of the employers who reported confidence in their employees' health care decisionmaking, only 38 percent provided regularly updated comprehensive information. Those professionals who are more apt to use the Internet in their professional or personal life or who routinely make financial and risk decisions are more attracted to CDHPs (Davis, 2004).

Plans with high cost sharing may pose additional challenges to providers in fee collection.

Physicians and hospitals may face billing and collection challenges in treating patients with HDHPs. Patients with HDHPs must meet a large deductible before their health plan starts to cover care; providers may experience considerable difficulty in collecting the billed charges from patients, leading to high levels of bad debt and high collection costs (Romano, 2006).

Back to top