Bar Journal - Spring 2007
Healthcare Pricing Revealed
By: Attorney Leslie Ludtke and Tyler Brannen
Few people understand how health care is priced, and even fewer know the actual price of a health care service or procedure. Until recently, the intricacies of the health care pricing system had little practical importance to the average person. Patients sought the care that their doctors believed that they should have. Health insurance paid for that care. Managed care employed a gatekeeper model, not a market-driven approach, as the primary mechanism for controlling costs. Insurance carriers negotiated contract prices with providers to establish the amount that the carrier would pay for certain procedures or services. When the contract price for health care services went up, and the cost of medical claims increased, insurance carriers raised the premiums to reflect the increased medical claims costs.
To contain the rising cost of the unit or contract price for medical goods and services, health insurance companies instituted a gatekeeper model to control utilization. Under this model, while the unit price for the service might increase, the increase in cost would be offset by reduced utilization of specialty health care services. Members in a managed care plan had minimal co-pays to encourage them to use preventive and primary interventions, rather than the services of specialists. The member cared little, if at all, about the actual price that the carrier paid for a service or procedure because once a carrier or a primary physician approved a service, the member had no financial liability beyond perhaps minimal co-pay.
During the early years of managed care, the gatekeeper model worked well in controlling rising medical claims costs. Gatekeepers, who were generally primary care physicians, controlled utilization of medical services through a tightly managed referral system monitored by the carrier. While effective initially, this model resulted in significant patient backlash. Patients demanded more choice and more expansive networks, and the health carriers responded by broadening their networks and relaxing the restrictions around referrals. As the gatekeeper model of managed care weakened, health care costs started to increase. At present, the rate of increase appears unsustainable. According to annual surveys conducted by the Kaiser Family Foundation, health care costs have increased nearly 90 percent since 2000 compared with increases in the consumer price index and workers’ earnings of slightly less than 20 percent.
With substantial increases in the cost of health care and increasing consumer resistance to the gatekeeper model for controlling utilization, the response has been to adopt a market-driven model to contain costs. Managed care products that previously did not require a financial contribution now are subject to high deductibles and co-insurance. Individuals, who formerly sought health care services as patients, have now become consumers of health care. Under this new health insurance paradigm, it is believed that shifting costs to individuals will unleash market forces stifled by the gatekeeper model and consumer demand will be allowed to function as a cost control mechanism.
From a practical standpoint, the efficacy of the market-based model in reducing the increase in health care costs is a matter of secondary concern to individuals. While patients or consumers of health care certainly hope that the escalating cost of health care services abates over time, their more immediate and pressing concern is how to pay for the health care services that they need right now. When both insured and uninsured individuals are expected to pay for an increasing portion of their health care costs, it is hardly surprising that they want to know how much the services will cost before they obtain them. Equally unsurprising is the new consumer demand for comparative cost information across different providers and carriers and comparative cost information on the price paid by the uninsured and the financially indigent.
I. New Hampshire’s Response to the Shift to a Market-Based Model of Health Care
A. The Creation of the Comprehensive Healthcare Information System
The New Hampshire Insurance Department (NHID) has taken a proactive approach in responding to the shift away from a gatekeeper model to a market-based model of health insurance. In 2003, the General Court established a Comprehensive Health Care Information System that is now codified in RSA 420-G:11-a. The Comprehensive Health Care Information System is a central repository of claims data and other health care data that allows the NHID to analyze and understand how the health care dollar is being spent in New Hampshire. RSA 420-G:11 requires all health carriers and third party administrators to electronically submit their encrypted claims data to the state.
The inclusion of all payer claims data in New Hampshire’s comprehensive health care information system provides the mechanism whereby the NHID is able to provide information to consumers of health care about the actual cost of health services in New Hampshire. In determining how to use the claims data, the NHID has been guided by the statutory statement of intent in the Comprehensive Healthcare Information System. RSA 420-G:11-a provides that the comprehensive health care information system shall function as a resource to make health care data available to insurers, employers, providers, purchasers of health care and state agencies.
New Hampshire is the only state in the nation that has created a comprehensive health care information system in this fashion, and one of only several states that collects claims data from both carriers and third party administrators. The claims database is essential both in determining the actual cost of health care services and in understanding what we, as a state, are purchasing with the money we spend on health care. It is the only window into how New Hampshire as a state spends its health care dollar.
The reason why the claims database is essential for understanding how health care services are priced is that health care services delivered to individuals covered by private insurance plans are paid at an “allowed” amount. The “allowed” amount is the negotiated payment level agreed to between the carrier and the provider and it is this amount that is used to determine the patient’s financial liability. Services that are covered benefits under a health insurance plan are paid at the “allowed” amount regardless of whether they are subject to a deductible or co-insurance. For example, if the charge is $2,000, but the contract rate is $1,000, the patient liability when based on a $500 deductible and a twenty percent co-insurance is $600. When the patient is insured, the provider’s charge rate is irrelevant when the health care service is a covered benefit.
The “allowed” amounts for services differ by carrier and provider. The cost for services provided to persons that access the health care system through governmental programs, such as Medicaid and Medicare, is an administrative rate set by the government. These rates differ depending on the government program and may be higher or lower than the “allowed” amount depending on the service. The cost of health care services provided to the uninsured is based on the charge rate, which is taken from the hospital or physician’s charge master. The charge master reflects the “list price” for the services. The list prices generally are not reflective of the cost of services, but rather are developed to produce a certain amount of revenue. Charge rates are often over 100 percent higher than negotiated discount rates or government rates. Most healthcare services are not actually paid for at the charge rate.
Although some states have addressed the need for price transparency by requiring the posting of charge masters, the charge master is not a useful tool for giving individuals an estimate of their financial liability because of its complexity and irrelevancy to insurance rates. Moreover, the charge master price does not reflect the actual cost of the service to an insured person, a person covered by a governmental program, or an uninsured person who receives a discount due to financial indigence or prompt payments. The charge master may be hundreds of pages long, is generally not cost-based, and often contains the charges for thousands of services and procedures. It is an administrative tool used by a provider and is not intended to provide the price information in a format that would allow an average person to determine the actual cost of a health care service.
The insurance claims data go beyond the charge master and allow for the development of information about how much is actually paid for health care. It is important to understand that there is no single price of a health service. The amount that an individual actually pays for a service will vary considerably based on how that individual accesses the health care system and where the service is obtained. Private insurance pricing, government pricing, and uninsured pricing differ considerably. Moreover, even when two different persons access health care through different insurance carriers, or different plans offered by the same carrier, the price may differ. It is not uncommon for two different insurance carriers to negotiate a substantially different price for the same service delivered by a provider, or for an insurance carrier to negotiate different payment amounts for the same service when offered in different types of plans, such as indemnity or managed care.
In short, the price that the individual or the carrier actually pays for a health care service or procedure will vary depending on a number of factors. These factors include: the provider that performs the service; the carrier that pays for the service if the person is insured; the individual’s insurance plan; the financial circumstances of an uninsured person; the health condition of the person obtaining the service; and differences in how the service is delivered and where it is delivered.
B. The Appointment of an Advisory Group
In 2004, the New Hampshire Insurance Commissioner convened an Advisory Group to assist with the development of a user-friendly and reliable information system that would support the shift to a market based approach to health insurance. The advisory group recommended that the information system be an internet-based support tool to give the health consumer access to the prices charged by different providers, the price actually paid by different carriers, and the amount charged to the uninsured for healthcare services.
Composed of representatives of insurance carriers and health care providers, as well as insurance producers, policy analysts, legislators, and employer and business representatives, the Advisory Group has undertaken several projects designed to provide more information to consumers of health care. In 2005, the department, with the assistance of the Advisory Group and the Center for Health Policy at the University of New Hampshire developed the first “Healthshop” website .
The first Healthshop website used a limited claims data set and displayed state wide price ranges and averages for certain types of health care services and procedures. The website contained a link to the University of New Hampshire’s Survey Center that was designed to elicit information from consumers and users as to their informational needs and suggestions for formatting and displaying that information. In the fall of 2005, the Advisory Group met to review the comments received from the Survey Center and determine how to redesign the website using the full claims database. In addition, the department, with input from the Advisory Group, initiated a study of the financial condition of New Hampshire’s hospitals, and the impact of free care provided to the uninsured on the finances of New Hampshire’s hospitals. This study is now available on the NHID website.
Many of the comments received in response to the initial Healthshop website requested comparative price information by carrier and provider. The redesigned HealthCost website responds to these comments by displaying comparative price information by provider based on the individual’s insurance plan and carrier. One column on the redesigned HealthCost website shows an estimate of the individual’s financial liability based on that person’s insurance plan; another shows an estimate of the plan’s liability, and the third shows comparative total cost by provider. For the uninsured, the website displays the median charge amount by the lead provider, and then calculates what the charge would be.
The claims data also show the substantial variability of costs that patients face with different providers. A column entitled “Precision of the Cost Estimate” provides a relative measure that reflects how consistent the charges were at the specific provider, for the procedure of interest, based on the claims experience. If most patients have a very similar cost outcome, then future patient experience is expected to more closely reflect the HealthCost reported amounts and the precision level is reported as ‘high.’
The HealthCost website also includes a measure of how the patient populations differ among providers. A final column, called “Patient Complexity,” reflects the relative illness burden of the patients served by different provider organizations. For example, patients with a recent history of congestive heart failure, osteoporosis, cerebral palsy, or cystic fibrosis would increase the patient complexity and theoretically the cost of care. Information on patient complexity is provided to offer insight about the health status of patients a provider may be more likely to attract, and provide information as to why the costs among different providers might differ.
C. The New HealthCost Website
The database that drives the HealthCost website is derived from the NHCHIS claims database. The database includes healthcare claims data from all carriers and third party administrators that either provide health insurance to individuals, small groups, and large groups or administer self-insured benefit plans that provide health coverage in New Hampshire. The database includes health insurance claims from all types of health care providers including hospitals, physicians, home health agencies, durable medical equipment providers, and ambulatory surgery centers. It also includes claims information from pharmacies and pharmacy benefit managers that provide coverage or administer prescription drug benefit plans.
D. The NHCHIS Database
The NHCHIS database is maintained through a Memorandum of Understanding entered into between the New Hampshire Insurance Department and the New Hampshire Department of Health and Human Services. The Department of Insurance is responsible for determining the format for collection of the claims data, and the Department of Health and Human Services is responsible for establishing the guidelines for release of the data. At present, the Maine Health Information Center (MHIC) is the contracted collection agent for the data under a contract with the Department of Health and Human Services. The MHIC is responsible for performing a series of edits and consolidation before the data is delivered to the NHID and the NHDHHS.
II. Claims Data to HealthCost Information
The data contained in NHCHIS are raw claims data (also known as administrative data), which the MHIC consolidates from NH carriers and third party administrators into a single dataset. In raw form, the data do not provide meaningful information to consumers about the cost of health care services. Several steps must be taken to convert the raw data that health carriers and third-party administrators file into meaningful information that a consumer can use to compare the price of health care services.
The first step is the initial data testing phase. This process looks for problems that may have downstream implications when analyses are performed. These tests include confirming that the volume of claim records and charges is relatively consistent for each month of the analysis; that the diagnosis and procedure codes are not truncated or missing; and that complete data sets have been submitted from each of the carriers and third party administrators.
Due to the volume of the claims, different data elements are stored in separate files. This second step requires the creation of a relational database, which allows files to contain minimal duplication of data fields that take up a lot of space. For example, one of the procedure codes for arthroscopic knee surgery is 29881. The definition that goes with this code might be: “arthroscopic knee surgery, meniscectomy, medial or lateral with meniscal shaving.” A relational database keeps the procedure code in one file that includes all claims history, and the description with the code in another file that just includes procedure descriptions. This avoids having records that repeat the definition thousands of times in the already large database that includes all procedures in the claims history. By including both the procedure code in the main claims database and in the “reference” database, the user can select just the claims with a code of 29881 and “join” that set of data with the description in the reference database. Even so, the databases when using NHCHIS can easily exceed 10 gigabytes, beyond the limitations of many popular software packages, such Microsoft Access or Excel.
The third step involves the aggregation of the data. As the NHCHIS claims data includes detailed information about the services and reimbursement to all types of providers, not just hospitals, it must be aggregated to allow a consumer to obtain comparative cost information. The NHID, with the assistance of the Advisory Council, developed a mechanism for aggregating charges to give consumers an estimate of the total cost of procedures that may include hospital, physician, drug, and equipment charges. As the patient may not be aware of all the different providers that may be involved in a service or procedure, each claim record that relates to that service is aggregated to develop a total cost for that procedure. A patient is unlikely to understand or have an interest in tracking payments made to different physicians involved in a single procedure, or in separating those payments from the payments made to the hospital or facility where the service is performed. For that reason, the HealthCost website displays a single total price.
To provide a “package price” for a specific procedure or visit, a series of steps are taken to build the total cost of the encounter. First, the procedure is identified by the procedure code. Then, all patient specific claims submitted by any provider that are related to the procedure are combined and charge and payment amounts are summarized. A “lead provider” is identified so that only one provider is shown in the HealthCost report page. The lead provider is the organization that is the most recognizable entity providing care. If care is provided in a hospital, then the hospital is listed as the lead provider. If care is provided by a physician practice, the lead provider is the name of group practice.
For example, when a birth delivery takes place, there are claims and payments for the mother and the newborn. The claims are from both the hospital and the physician and payments usually do not go to the same place. In HealthCost, all of these claims are combined and assigned to the hospital as the lead provider. This approach differs from that used on other websites that report charges only, and separate the charges for the mother and newborn. Generally, on other websites, physician charges and payments are not reported. The example below illustrates the cost information that would be shown on the HealthCost website. The charge column will not be included and is shown here only to distinguish the charges from the “allowed amount,” represented here by the “Estimate of Combined Payments.” The estimates are based on the Anthem insurance carrier, for the PPO insurance type, with a $500 deductible and 20 percent co-insurance.
For individuals interested in the specific cost of the components of a service, such as lab work and radiology, rather than the cost of the entire visit, HealthCost will provide information on the charge and paid amounts for specific common procedures. This information will reflect the average amounts for all providers and insurance carriers in the database, rather than detailed provider and carrier information. There will also be separate columns for the amounts attributable to the hospital and to the physician. Since this type of data is less useful to typical consumers, the website will not automatically provide this report through the normal process of obtaining HealthCost information. Instead, it will be available through a link as supplemental data once the procedure of interest has been reported on.
A. Identification of Services
The department, with the assistance of the advisory council, developed a list of health care services for inclusion in the HealthCost website. In selecting the list, the department considered procedures or services that were common and that would be available from many different providers. The list is not intended to include all procedures or services. It is expected that additional procedures will be added in response to consumer demand for more information. At present, the HealthCost website includes price information on: emergency visits, arthro-scopic knee, surgery, colonoscopy, hernia repair, gall bladder removal, tonsillectomy with adenoidectomy, breast biopsy, kidney stone removal, mammogram, vaginal and c-section, birth and newborn, computed tomography scans (CT scans), x-rays, ultrasounds, MRIs, physicals, and lab tests.
B. Calculation of “Cost”
“Cost” is a term used loosely in health care. In HealthCost, what is reported as cost is a calculated estimate derived from multiple patient experiences for the same health care service from the same provider. Because the website is designed as a tool for consumers, the amount paid by the patient is the cost of primary interest. To estimate what the patient will pay, several specific components are considered. The health care insurance carrier and specific health insurance coverage provided will determine the price paid for the health care service. In addition, the amount of the insurance deductible and coinsurance will impact cost. All these factors are included in the website. Information that may affect how much an individual may pay for a service is requested from the user.
The claims data are analyzed based on the experience of multiple carriers at a particular provider, but the cost estimate is specific to the individual’s carrier and coverage. Because costs are aggregated and the process of providing care is highly variable, the calculated median total cost is analyzed within context of overall utilization at the provider. Therefore, conclusions can be drawn with the advantage of a larger sample size.
The median is reported instead of the average. Reporting the median accomplishes two objectives. Consistent with the purpose of HealthCost, the median provides a better measure of central tendency when predicting what an individual patient will face when obtaining care from the provider(s). The median is influenced less than the average by outlier observations that can skew the results. The second objective, as illustrated in the following example, is that the median makes it more difficult to determine actual contract terms for payments between the insurer and the provider.
In this example, both insurance carriers would have the same median reported in HealthCost:
Because median is reported instead of average, the actual contract allowed rate cannot be determined. Based on the median, the contracts appear identical. The average would be a more accurate representation of the “value” of the contract to the insurer and the provider. However, $100 provides a good estimate of the total cost for most patients, regardless of which insurance carrier covers them.
C. Data Variability
Health care cost data are highly variable. When rates are reported in HealthCost, the NHID will include information on the variability of the data. If the historical data show low variability (total costs are consistently very similar for a given procedure), then the precision of the cost estimate is considered “high.” Likewise, if the data show extensive variation, the estimate will indicate that individual experience is more likely to deviate from the reported rate.
The measure of variation in the rate is based on a statistical measure called the “coefficient of variation” which reflects the difference between the median charge for the insurance company product line and the overall median for all insurance companies and product lines at the provider identified. These values, both percentages, are summed together and translated into an ordinal scale. Like most ordinal scales, the distinction between the values at neighboring points on the scale are not necessarily the same. The scale is determined based on how the variability compares to other reported insurance carrier calculations within the health care service selected. The breakdown is based on percentiles, using 75th, 50th, and 25th break points.
D. Risk Adjustment
Risk adjustment is a way of describing how populations differ from one another for the purposes of assessing the resources needed to care for them over a period of time. Risk adjustment is provided in HealthCost to help explain why the historical cost to patients at one provider may exceed that at another provider. Risk adjustment considers more than the diagnoses for the visit of interest. Instead, all of the diagnoses throughout a given year are considered so that the effect of multiple disease conditions can be considered in evaluating how one patient population differs from another. Patient populations that average a more significant disease burden are expected to need greater health care resources than a less complex patient population.
The application of risk adjustment is specific for the patients and the identified condition. For example, Hospital A attracts a very “average” patient population when all treatments are considered, but Hospital A attracts very complex patients for normal vaginal deliveries. When viewing the cost rates for deliveries, the patient complexity at Hospital A would be described as “high.”
“Behind the scenes,” the risk adjustment calculation is a relative index measure, where 1.0 is the mid point, and values above or below are a calculated difference in expected resource consumption. For the HealthCost website, the index measure is translated to an ordinal scale based on the index value when compared to other reported insurance carrier calculations within the health care service selected. The breakdown is based on percentiles, using the 90th, 75th, 25th, and 10th separation points.
The rates provided in HealthCost are not risk adjusted. They are the actual calculated rates based on the data and the algorithms. The risk adjustment field is provided to offer one explanation why the costs shown may be greater than or less than that of another provider.
A process exists to remove outliers, which are values that are unusually high or low. Outliers are data values that do not represent the typical experience for a particular service at a particular provider location, and they can exist for several reasons. First, claims data are not perfect. In some cases the claims are incomplete. These circumstances may exist when the providers have not billed for all services yet, or the insurance carrier has not processed all of the claims submitted for the visit. Alternatively, human error may result in a particular service that is coded incorrectly. An extreme example might be a service related to a kidney transplant that is coded as a kidney stone removal. In this example the cost for the kidney stone removal would appear to be extremely excessive. Because the median is calculated instead of the average, outliers have a small effect on the estimated costs reported in HealthCost, but they can have a substantial impact in the formula used to assess the variability in the rates.
Removal of the outliers takes place at two points. First, a ceiling for total charges to be considered in the analyses is established. The ceiling is identified by assessing where 95 percent of all charges fall below, across all providers. In many cases, the values above this threshold are for one or two providers that are performing a service atypical of what is normally performed elsewhere. Therefore, the experience is considered too extreme to compare to other providers in the analysis, even with the consideration of risk adjustment.
The second point where outliers are removed is after analyzing a specific provider’s experience. Patients with total charges in the lowest fifth percentile or highest fifth percentile are removed from the analysis. The calculations of the percentiles is done using standard statistical conventions, so if the observation values to do not vary much from each other, it is unlikely any will be removed.
F. Price Transparency
Price transparency is a common term in health care today. One of the primary drivers of this interest is the expectation that patients need to pay more of the bill, and without transparency, the bill is a mystery until several weeks after the patient has been treated. Even with transparency, the methods of delivering health care and deriving costs are complex and evolving in such a way that they pose substantial challenges for efforts such as HealthCost. Despite an extensive history of claims to draw information from, an individual’s particular experience may be quite different from the average patient, resulting in costs that are substantially more than what is shown on the HealthCost website.
The HealthCost effort at price transparency is more than about unit costs. Unit costs means the price paid for each item that is billed during a visit of care. Due to the bundling nature of the methodology used in HealthCost, the “price” is as much about utilization as it is unit cost. A provider with low relative unit costs may still appear expensive if that provider typically orders twice as many lab tests or twice as many consults from various physicians.
Transparency, applied to more than just health care costs, is an evolving concept that will allow many traditional “unknowns” in health care to be evaluated. The NHCHIS allows researchers to begin to understand the patterns that exist in the way health care is delivered and why Americans pay more and receive less than most other industrialized nations.
III. HealthCost for Employers
The first phase of HealthCost focuses on the individual patient as a consumer. As employers make many of the decisions related to purchasing health insurance, the NHID will develop a HealthCost portal for employers. The goal of this initiative is to use current data sources to provide information that employers can use to select health insurance carriers and benefit products, determine the group’s cost drivers, evaluate purchasing alliances, and understand the factors that influence the cost and delivery of health care.
The primary data source for the patient-focused initiative is the NHCHIS claims data. When developing the employer section, multiple data resources will be used, including the NHCHIS database; the discharge data from NH hospitals Uniform Hospital Discharge Data Set (UHDDS); the Health Plan Employer Data and Information Set (HEDIS), which includes 57 separate measures organized into five health topic areas; and the NHID Supplemental Report, a report on the distribution of insurance based on geographic, market segment, and benefit designs. These data sources will allow employers to review the distribution of health insurance products sold in NH based on the line of business (HMO, PPO, POS, and Indemnity), employer group size, loss ratios, basic coverage categories, and levels of cost sharing, such as deductibles and co-insurance. Information will also show relative medical claims cost differences by community, with adjustments based on variables such as age, gender, and health status. Most of the data can be displayed by specific geographic areas of the state. The employer portal will be developed throughout the current year, with release planned by January 1, 2008.
The health insurance system continues to evolve and place more financial responsibility on the member as a patient and consumer. This change creates a strong demand for new information that can assist consumers in choosing where to receive care and what it will cost them. New Hampshire is one of the first states to develop an information system where health care cost data can be compiled and translated. These efforts are key steps in the development of a system based on price transparency and consumer-directed health care.