In most standard economic evaluations, costs reflect the total resources utilized in the output produced; that is, the manufacture of goods or the provision of services provided. This is not usually the case in health care, nor in palliative and end-of-life care, because it is unclear what constitutes the units of service or output. With the exception of component product-services such as laboratory or radiological testing, or discrete services such as outpatient procedures, the provision of health care consists of a range of services. Moreover, as much as it has been attempted, these services are not directly or easily translated into one type of output, such as improved health status, alleviation of illness and symptoms, or increases in quality adjusted life years.
As a result, existing accounting systems do not identify and track many resources and component costs of service delivery (i.e., wages and salaries, number of personnel needed to deliver a "unit of health care service," costs of plant and depreciation) that might be allocated to the end products or services. Some large insurers have tried to overcome the lack of connection between resources used and health care services provided. Medicare has developed "approximate" payment methodologies that "bundle" unaccounted-for costs into a set of payment categories from which health care providers are paid (e.g., DRGs - Diagnosis Related Groups - and managed care reimbursement formulas such as per diems). Managed care plans have relied on the experience-rated data on an insured pool's consumption of health care services as the basis for determining "cost-based" monthly premiums. The net result for most health care systems is that cost-related data from many health care accounting systems, such as billing charges and insurance claims, do not reflect the component costs of resources used in providing health care services.
The paucity of established accounting methodology to link raw materials and resources to the provision of health care services severely challenges development of health care service units to use as building blocks for calculating costs of care. The level of specificity for service units varies by program. For example, "hospice" represents a heterogeneous basket of different service units, and the level of detail in identifying service units within the "hospice" line item often varies in accordance with the perspective of the researchers and goals of the study. This is inevitable and appropriate. However, it is essential to clearly define the "units" and explain the advantages and limitations as appropriate. For example, in some studies it will be sufficient to quantify days of hospice care as a relevant cost unit. For others, components of hospice services, such as the number of visits and time spent by hospice nurse, social worker, physician and chaplain, will be required. Tabulating the "costs" of a service will depend on what the research questions are, what sources of data are available and the capacity of the research plan and budget.
Cost measures constructed from a service unit level have potential for avoiding many of the distortions and problems associated with idiosyncratic institutional-based cost accounting systems. The service unit approach requires a value standard for conversion of service units into dollar costs. Prices of health care service units that incorporate costs of component materials and resources, as well as the "opportunity costs" of those units provided would be ideal. Unfortunately, health care services are not provided through well-organized, competitive markets where supply and demand set per unit prices and enable opportunity costs to be accurately determined.
In addition to the previously noted problems of defining final services and output, health care is not a typical economic "good" or "service," to which value can be assigned by market forces. Consumers usually do not have a "choice" in whether they become sick and need health care services. They usually rely on experts and the information or recommendations of providers in choosing their health care "needs." As a result, market prices are not readily available or meaningful for valuing health care service units.
What are readily available are health care provider costs and revenue data, although they are not good measures of value. Value measures based on dollar values of billed charges, health plan and health insurance claims data and other administrative data sources include a number of non-service elements that make poor proxies for converting service units into costs. Cost-shifting between patient groups based on third party payer rates, institutional variation in price markups and profits, and inclusion of non-revenue cost centers and bad debt in many accounting systems obscure the specific economic costs of care. These factors apply whether one looks retrospectively at the costs during the last year of life or, prospectively, at the costs of palliative care for patients with limited life expectancy. Accounting data include non-resource charges and are sometimes available at considerable lag after the consumption of health care.
The monetary value (price/cost) of service units should be determined.
Monetary values or costs per unit of service units may come from Medicare
or other government payers, commercial payers, surveys of providers,
surveys of employers or employees (wages), queries of patients, published
studies and some combination of methods. The objective is to assign monetary
values to all resources and materials used that reflect their opportunity
costs from the cost perspective identified.
For example, a cost study from a public health perspective might use the Medicare reimbursement system, which was designed to approximate the cost base for health care services as determined by local area wage and price levels, historical utilization and institutional status (e.g., teaching versus community hospital) as the common and acceptable set of cost formulas. A cost perspective of private health insurers would require price and economic cost data from private sector health plans as well as actuarial statistics for some population of health care consumers.
Costs of care for each patient or group of patients should be determined by multiplying the total units of a particular service utilized by its monetary value to generate a per-unit cost and summing costs for all units of service utilized.
The use of health care service units as basic building blocks for cost accounting form a major component of the Workgroup's recommended standards and guidelines. An accurate cost per unit of services would allow comparisons of palliative and end-of-life care services between individuals with different demographic characteristics (e.g., children versus the elderly), programmatic models (hospice versus home health) and over different time periods (five years ago versus two years ago).
A cost accounting template is included as Appendix B (PDF 410 KB / 10 pp.) that captures the major medical services involved with palliative and end-of-life care programs. As presented by the Workgroup, the template indicates the details being collected at individual sites. For purposes of this report, the columns have been modified to indicate resources that might be captured at three levels of detail-simple counts, adjusted counts and billing units.
The M.C. Escher piece at the top of this page
is titled "Path
<<< Previous Next >>> [ Go Up ]
Promoting Excellence in End-of-Life Care was a national program of the Robert Wood Johnson Foundation dedicated to long-term changes in health care institutions to substantially improve care for dying people and their families. Visit PromotingExcellence.org for more resources.