Dr. Gerard Anderson
Professor, Department of Health Policy & Management and International Health, Bloomberg School of Public Health
Johns Hopkins School of Medicine
A Review of Hospital Billing and Collection Practices
House Energy and Commerce Committee
Subcommittee on Oversight and Investigations
June 24, 2004
2123 Rayburn House Office Building
Mr. Chairman, members of the Committee; my name is Dr. Gerard Anderson. I have been working on hospital payment issues for many years. Between 1978 and 1983, I worked in the Office of the Secretary in the US Department of Health and Human Services. In 1983, I was one of the primary architects of the Medicare Prospective Payment legislation. Following passage of the Medicare Prospective Payment legislation, I joined the faculty at Johns Hopkins where I have been for the past 21 years. At Johns Hopkins, I direct the Johns Hopkins Center for Hospital Finance and Management - the only academically based research center focusing exclusively on hospitals. I am also a professor of Health Policy and Management and professor of International Health in the Bloomberg School of Public Health and Professor of Medicine in the School of Medicine at Johns Hopkins University.
I would like to begin my testimony by highlighting several milestones in hospital payment policy. Because of the evolution of hospital payment policy, self pay patients are currently being charged 2 to 4 times what people with health insurance coverage pay for hospital services. These are not market rates and need to be lower. After reviewing the milestones, I will then make a series of specific suggestions to the committee that will make the current hospital payment system more equitable to the self pay patients. My preferred option is that hospitals be limited to what Medicare pays plus 25 percent.
Critical Milestones That Have Led To Market Failure in Hospital Payment
One hundred years ago most hospital care was either free or very inexpensive. In 1900, hospitals could provide little clinical benefit for most illnesses and were primarily places for housing the poor and insane who were sick. Hospitals were primarily philanthropic organizations. They were established primarily in poor urban areas.
Beginning in the 1920s, the ability of hospitals to improve the health status of patients increased dramatically. For the first time, rich and poor Americans sought out hospital care when they became seriously ill. Anesthesia expanded access to surgery and antibiotics made it easier to treat infections.
Physicians had a wider range of services to provide to hospitalized patients. New drugs and new equipment became available and better and more highly trained personnel were required to provide these services. The cost of providing hospital care began to accelerate. In order to recover these higher costs, hospitals began to charge patients for services. Hospitals developed a charge master file. Initially there were only a few items on the list. It listed specific charges for each service the hospital provided. A hospital day had one charge, an hour in the operating room had another charge, and x-ray had a third charge, etc. As the number of services the hospital offered increased, so did the length of the charge master file. There are now over 10,000 items on most hospital charge master files.
Before 1929, there was no health insurance and patients paid the hospital directly. In 1929, Baylor Hospital in Dallas, Texas began a program selling health insurance to school teachers in the Dallas County School district. Baylor created this health insurance system because many of its patients were having difficulty paying hospital bills. It became the prototype Blue Cross Plan. As the depression worsened in the 1930s, the ability of people to pay their hospital bills also worsened. Blue Cross and other types of insurance programs proliferated. These insurers paid charges based upon the charge master file.
During this period, the charges were based on the cost of providing care plus a small allowance for reserves. The markup over costs was typically less than 10%.
Private health insurance received a major boost during World War II when Congress made health insurance tax exempt. After World War II, private insurers continued to pay the charges that hospitals had established. Over time, the ability of hospitals to improve the health status of their patients increased, the kinds of services provided by hospitals increased and the costs of hospital care began increasing at 2 to 3 times the rate of inflation. By 1960, the typical hospital had established a list of prices for approximately 5,000 separate items. There were no discounts; everyone paid the same rates. The rates that insured and self pay people paid were similar.
Hospitals set their prices for these 5,000 items on a few criteria. The most important factor was costs. Charges were typically set at a given markup over costs, usually 10 percent. The hospital would estimate how much it cost to deliver a service and then charge 10% more. The ability of hospitals to estimate cost for individual services, however, was extremely limited by cost accounting. No hospital really knew how much it costs to provide a particular service because cost accounting techniques were not sufficiently detailed.
Market forces determined charges for only a few services. Child birth for example, was one service for which patients could engage in comparative shopping. Pregnant women had almost nine months advance warning that they would be admitted to the hospital and their families could therefore engage in comparative shopping. In theory, they could compare differences in the out-of-pocket costs and the perceived quality between two hospital delivery rooms. Thus, hospitals kept delivery room charges at or below actual costs.
For most services, however, it was often impossible for consumers to engage in comparative shopping because either the admission was an emergency or their doctor had admitting privileges in only one hospital. For most admissions, they had no idea what services they would use during their hospital stay. They could not engage in comparative shopping if they did not know what services they were going to need. In addition, for most people, insurance paid the full bill and so patients had no financial incentive to engage in comparative shopping.
Medicare Becomes Involved
When the Medicare program was established in 1965, Congress decided that the Medicare program would pay hospital costs and not charges. This was the method of payment used primarily by Blue Cross. Congress recognized that charges were greater than costs and that the Medicare program would be able to exert little control over charges. A very detailed hospital accounting form called the Medicare Cost Report, was created to determine Medicare’s allowable costs.
In order to allocate costs between the Medicare program and other payors, the Medicare program required hospitals to collect uniform charge information. Uniform charges were necessary in order to allocate costs to the Medicare program. The Medicare Cost Report could determine allowable costs for the entire hospital, however, it needed a way to allocate these costs specifically to the Medicare program. Charges are used to allocate costs to the Medicare program. If, for example, 40% of the charges were attributed to the Medicare program, then the cost accounting system would allocate 40% of the costs to the Medicare program.
In order to prevent fraud and abuse, the Medicare program required hospitals to establish a uniform set of charges that would apply to everyone. Otherwise, the hospital could allocate charges in such a way that would result in more costs to the Medicare program.
Hospitals continued to have complete discretion on how they established their charges. The Medicare program did not interfere with how hospitals set charges for specific services. One hospital could charge $5 for an x-ray and another hospital $25 for the same x-ray. A number of studies conducted at the time showed wide variation in hospital charges.
People with insurance generally had little reason to scrutinize their bills because they had first dollar coverage. Insurance paid the full hospital bill. Also, patients did not know what services they would need and so they did not know what prices to compare. Insurance companies did little to negotiate with hospitals regarding hospital charges in the 1960s and the Medicare and Medicaid programs did not pay on the basis of charges.
In the 1970s, market forces still had a small impact on hospital charges. In reality, the hospital had virtual carte blanche to set the charges. The number of separate items that had a charge associated with them, doubled from 5 to 10,000 at the typical hospital, where it is today.
Two major changes occurred in the 1980s that had a major impact on hospital charges. First, Medicare created the Prospective Payment System which eliminated any need for using hospital charges to allocate hospital costs. Second, most insurers began negotiating discounts off of charges or using some other mechanism to pay hospitals. As a result, any market forces that existed to limit what hospitals could charge were almost completely eliminated.
In 1983, the Medicare program moved away from paying costs and instituted the Prospective Payment System (DRGs). As the Medicare Prospective Payment System became operational, the need for the Medicare Cost Report and therefore the need for a uniform charge master file to allocate costs became less and less important. Today, because nearly all of the Medicare program uses some form of prospective payment, the requirement of a uniform charge master file by the Medicare program is virtually unnecessary.
Managed care plans began to negotiate with hospitals in the early 1980s. They wanted discounts off of charges in return for placing the hospital in their network. They successfully negotiated sizeable discounts with hospitals. As insurers began to compete with managed care plans in the mid 1980s, they also began to move away from paying full charges and started negotiating their own deals. Some insurers decided to pay on a per day basis, others decided to pay discounted charges, or a negotiated rate. Nearly all private insurers and managed care plans stopped using full charges as the basis of payment by 1990. They simply could not compete in the market place if they paid full charges.
Cost Shifting and Market Failure
As each segment of the market developed a different way to pay hospitals, this lead to a phenomenon known as “cost shifting”. As the Medicare program instituted the Prospective Payment System (DRGs), the Medicare program began to limit the amount that Medicare would spend. Faced with constraints on Medicare (and soon thereafter Medicaid) spending, the hospitals began to engage in “cost shifting”.
To do this the hospital industry increased prices to commercial insurers. Given that most commercial contracts were written to reimburse hospitals based on the hospital’s own charges, it was relatively simple matter for hospitals to raise their prices. When commercial insurers tried to raise prices to the employers, however, employers began to examine alternatives. Employers slowly and then rapidly embraced managed care. Managed care expanded rapidly using their market power to negotiate discounts off of charges with hospitals. Soon commercial insurers asked for similar discounts. Private insurers continued to pay more than Medicare however in most cases.
Without the federal government, state governments, private insurers, or managed care plans paying full charges, the regulatory and market constraints on hospital charges were virtually eliminated. By 1990, the only people paying full charges were the millions of Americans without insurance, a few international visitors and the few people with health savings accounts. These individuals had limited bargaining power and were asked to pay ever increasing prices. Effectively, there was market failure in this aspect of the hospital market.
Without any market constraints, charges began increasing much faster than costs. In the mid 1980s charges were typically 25% above costs. Without any market constraints, it is now common for charges to be two to four times higher than costs. Charges are also two to four times what most insurers pay. Most insurers, including Medicaid, Medicare, and private payors, pay costs plus/minus 15 percent. Over the past twenty years, the difference between what the hospital charges and what it costs to provide care has grown steadily in nearly all hospitals.
Hospitals have been able to increase charges because self pay individuals have limited bargaining power when they enter a hospital. They first must find a team of physicians willing to treat them who also have privileges at that hospital. Then they must negotiate with the hospital. Often they wait until they are ill before they seek medical care. This further diminishes their bargaining power because it is now an emergency. Often the hospital wants prepayment. Because most self pay persons have limited resources and cannot make full payment in advance, this further diminishes their bargaining power.
Perhaps the most important constraint on their bargaining power, however, is that they do not know what services they will ultimately need. They do not know how long they will remain in the hospital, what x-rays or lab tests they will need, and therefore they cannot know in advance what services they will require and which of the 10,000 prices they should negotiate.
Costs, and What Insurers Pay in Pennsylvania
Using the most recent data available I compared what insurers pay and what hospitals charge in Pennsylvania. As noted earlier, charges vary considerably from hospital to hospital. Pennsylvania collects data on what hospitals charge and what insurers pay in Pennsylvania for different illnesses (www.phc4.org). For example, I looked at the charges that Philadelphia area hospitals charged for medical management of a heart attack in 2002. The average charge was over $30,000. Most insurers paid less than $10,000.
Why Are Charges So Much Higher Than What Insurers Pay?
There are three main reasons why hospitals set charges 2-4 times what they expect to collect from insurers and managed care plans. The first is that Medicare outlier payments are partially based on charges. The second is that bad debt and charity care is typically calculated at full charges. The third is that some self pay patients actually pay full charges.
In the Medicare program, a small proportion of patients are much more expensive than the average patient. These are known as outlier patients. Medicare pays for these patients outside of the DRG system. Medicare continues to use charges as part of the formula used to determine outlier payments.
Recent investigations have shown certain hospital systems manipulating the payment system in inappropriate ways to over charge the Medicare program for outlier patients. One aspect of this fraud was the exceptionally high amounts these hospitals charged. Lowering the charges would diminish the over charges in the Medicare program for outlier payments and would reduce the level of fraud.
Second, hospitals routinely quantify the amount of bad debt and charity care they provide. This helps with fund raising and is used to meet charitable obligations. However, by valuing bad debt and charity care at full charges, these numbers vastly over estimate the amount of bad debt and charity care the hospital actually provides.
There are three groups that still pay charges. The first are people who have health savings accounts. Some of these individuals may be able to negotiate discounts although most pay full charges. It is extremely difficult for one person to negotiate with a hospital, especially in an emergency situation. The hospital holds all of the cards. Lowering the charges will benefit people with health savings accounts.
The second category is international visitors. These are typically affluent individuals who need a procedure that can be performed most effectively in the United States. These individuals are willing to pay full charges, even at inflated prices.
There are compelling arguments to charge international visitors higher prices than Americans. Most can afford to pay and, in addition, they have not subsidized the hospital sector in the United States through tax payments and other public subsidies. On the other hand, in most other countries Americans are usually treated free of charges if they have an emergency. An American injured while traveling in Canada, Australia, France, etc would be treated free of charge or receive a very small bill. Although there is no data that I know of that would allow us to compare the cost of care provided to Americans traveling abroad to the cost of care provided to foreigners receiving care in the U.S., I expect it would be similar. In that case it seems unfair to charge foreign visitors so much more for a service when Americans receive care free of charge overseas.
Impact On The Uninsured
The third, and by far the largest group that is asked to pay full charges is the uninsured. There are 43 million Americans who are uninsured. The uninsured can theoretically negotiate with hospitals over charges, but they have little bargaining power. My review of hospital practices suggests that less than 1 in 20 uninsured patients actually negotiates a lower rate.
Many uninsured people are unable to pay full charges. In fact, most studies suggest that less than 1 in 10 uninsured people pay a portion of their charges and relatively few pay full charges. In fact, in most hospitals only 3 percent of total revenues comes from people who are uninsured. Self pay patients represent a very small proportion of hospital revenues.
The toll on the uninsured, however, can be substantial. There are numerous reports that show hospitals attempting to collect payments from the uninsured. The people who do not pay are sent to collection agencies and some are driven to bankruptcy. One study found that nearly half of all personal bankruptcies were related to medical bills (M.B. Jacoby, T.A. Sullivan, E. Warren, “Rethinking The Debates Over Health Care Financing: Evidence from the Bankruptcy Courts,” NYU Law Review 76, May 2001: 375). Another survey (D. Gurewich, R. Seifert, J Pottas, The Consequences of Medical Debt: Evidence From Three Communities, The Access Project, February 2003) found that hospitals were routinely requiring up front payments, refusing to provide care, or encouraging uninsured patients to seek new providers if they did not have health insurance. Many respondents found the terms the hospitals were offering were difficult to maintain given the hospitals’ inflexible collection processes and their own financial situations.
Nearly all hospitals do this to some extent. For example, a series of stories in the Wall Street Journal examined the collection procedures at Yale- New Haven hospital. The Wall Street Journal found that in 2002, the Yale- New Haven hospital was lead plaintiff in 426 civil lawsuits, almost all of which concerned collections or foreclosure lawsuits against individuals, compared with 93 lawsuits at a similarly sized local hospital. Yale- New Haven Hospital also frequently engaged in aggressive collections measures, such as wage garnishment, seizure of bank accounts, and property liens. In 2001, the hospital filed 134 new property liens in New Haven, almost 20 times the number filed by the city’s other hospital.
Benefits of Lower Charges
If charges were lowered there could be two beneficial outcomes. First and most important, fewer self pay individuals would declare bankruptcy. Second, more self pay patients would be able to pay their bills if the charges were more in line with prevailing rates.
Guiding Principles for Setting Rates
The question therefore becomes what is a reasonable rate for hospitals to charge self pay patients given that neither market forces or regulations constrain hospital charges.
I propose four guiding principles. First, the rate should not interfere with the market place. The rate that self pay individuals should pay should be greater than what insurers and managed care plans are currently paying hospitals. Second, the charges should not be substantially higher than what insurers and managed care plans are currently paying hospitals. Individuals with limited bargaining power should not be asked to pay exorbitantly high rates because they lack market power. Third, the rate should be transparent to patients. Patients should know the prices they will be asked to pay when they enter the hospital. Fourth, the system should be easy to administer and to monitor.
Two Payment Alternatives
I have two specific suggestions for the Congress to consider.
The first is to mandate that the maximum a patient can pay is the amount paid by Medicare plus 25%. I call this DRG+25%. The rationale for allowing hospitals to charge 25 percent more than Medicare is based on three factors. First, private pay insurers pay an average of 14 percent more than Medicare for a similar patient. I then add one percent for prompt payment. Finally, an additional amount (10%) is added because the amount paid by private insurers is an average and some commercial insurers pay more than the average. Adding the three factors together results in a proposed payment rate of DRG + 25%.
The advantages are that the DRG + 25% rate is easily monitored and adjusts for complexity of the patient. It would be continually updated by Medicare as Medicare updates the PPS rates. The disadvantage is that the rate is not market determined. In most markets, however, it would be above what insurers and managed care plans are paying.
A second option is to allow hospitals to charge the maximum they charge any insurer or managed care plan on a per day basis. The advantage is that it is market determined.
There are four disadvantages. First, it will require regulations and auditing to verify the rate is the maximum they charge any insurer or managed care plan. Second, in order to make the rate transparent, it will be necessary to keep the rate in place for an extended period of time, probably a year. This interferes with the market place. Third, it will require hospitals to tell all insurers and managed care plans who was the worst negotiator. This also interferes with the market place. Fourth, it requires all negotiations to be on a per day basis. Any other payment system would be too complicated. This interferes with the market place.
Balancing the pros and cons of both options, I recommend the DRG+25% option. It complies with all four principles- it is above what insurers are paying, it is a reasonable amount, it is transparent, and it is easy to monitor and verify.
Rate Is Too Low
Insurers may argue that they are entitled to more substantial discounts over self pay individuals for two reasons- prompt payment and volume discounts. The prompt payment argument has some validity. A two month delay in payment at a 6 percent interest rate is equivalent to a 1 percent savings. This is built into the DRG + 25% payment.
The volume discount argument is more complicated. In my opinion it has limited financial impact, especially on medical services. Most insurers and managed care plans do not guarantee a certain volume of patients and certainly they do not guarantee a certain case mix of patients. Instead, they agree to put the hospital on a preferred list of hospitals. The patient and the physician still make the final decision regarding which hospital to select. The choice, therefore is fundamentally different from a purchase in the manufacturing or retail sector where a large volume of goods or services is actually purchased.
The second part of the volume argument, however, is probably more important. The same medical services will be used if the patient is self pay or insured. The patient will use the same set of laboratory tests, spend the same time on the operating table, require the same nursing hours, etc. The medical services are what is most expensive in a hospital and this does not depend on the volume of patients that an insurer has.
Incentives to Purchase Health Insurance
Some individuals with high incomes choose to self insure. An important and difficult question is whether these individuals should be able to get the benefits from these lower rates.
One argument is that these individuals have voluntarily chosen to go without health insurance and they should pay a much higher rate if they get sick. A second argument is that these individuals should be given financial incentives to purchase health insurance and that lowering the hospital rates for them will only induce them to go without coverage.
Although there is merit in both arguments, the question is what is a fair rate for them to pay when they get sick? When they need hospitalization they should pay a rate that is somewhat higher than people with health insurance coverage pay. The DRG +25% criterion meets this objective. This group of people should not be asked to pay for the bad debts of other self pay patients any more than the insured population. And, if the rates were reasonable they would be more likely to pay.
Simplification of Payment System
The medical care system could be simplified if such a change were enacted. One major change would be the elimination of the Medicare Cost Report. A second simplification is that it would be easier to calculate any discounts that hospitals are offering to low income individuals.
The Medicare Cost Report was created in 1965 with the passage of the Medicare legislation and the decision by the Congress to pay costs. The Medicare cost report is now a document that is over 6 inches thick and requires many hours for hospitals to complete. However, with the passage of the Medicare Prospective Payment legislation in 1983 and subsequent adoption of additional Prospective Payment Systems for outpatient care etc., there is no longer a compelling reason for maintaining the Medicare Cost Report. Any information the Congress needs from hospitals to set hospital payment rates could be summarized in a few pages. The only relevant information is the profit of hospitals and some information used to calculate graduate medical education and disproportionate share payments.
Hospitals often give discounts to low income self pay patients. It is therefore key to understand what is the basis for the discount. A discount from full charges is not really a discount if it is still greater than what insurers and managed care plans would pay. A true discount would be below what public and private payors are expected to pay. If the payment system for self pay patients were simplified (DRG + 25%) then it would be easier for them to determine if they are really getting a discount and how much they were expected to pay. Currently the self pay person does not know the real extent of the discount or how much they will pay.
In summary, what should be done?
1. Both Congress and the hospital industry should recognize that hospital charges are not determined by market forces. The only people paying full charges are those with limited or no bargaining power.
2. The maximum that self pay individuals should have to pay for hospital services should be DRG rate plus 25%.
I would be happy to answer any questions.