Utilization management in Alberta's new funding environment.

Wetaskiwin Health Care Centre has a long tradition of caring for the ill in its agricultural community. In July 1991, a new $41.5 million facility was opened, with 135 acute care and 80 extended care beds. The Hospital District also operates another 50-bed extended care facility. It has an annual operating budget of $18 million and about 550 employees.

In the mid-1980s, Wetaskiwin Hospital was selected as the Alberta test site for the implementation of new national guidelines2 for the collection and reporting of resource utilization and costing information. These guidelines provide a method for measuring the resources used in treating patients at both the functional center (departmental) and the patient-specific levels.

Although the hospital was successful in costing the services provided to individual patients and groups of patients, it was difficult to use this information for utilization management without comparative data from other hospitals, or without clinical profiles of ideal resource utilization and costs. This changed, however, with the implementation of a new funding system in the province.

The Alberta Acute Care Funding System

In December 1988, Alberta Health initiated its Acute Care Funding Plan in response to increasing concern about the existing method of funding acute care hospitals. New methodologies were developed to fundamentally change the way hospitals are reimbursed. Key objectives were to:

* Improve fairness and equity in the fund in, of hospitals.

* Provide incentives for cost-effectivenes and shared services.

* Move to an output (case-based) funding scheme, rather than the continue funding of inputs (such as staffing an supplies).

* Use clinically meaningful diagnostic groups.

* Recognize the differing severities of patients.

Although Alberta Health is developing funding schemes for day surgery, emergency care, and other ambulatory care services, the following overview focuses on the inpatient component of the Acute C Funding Plan.

Severity-based Case Mix Grouping

Because an objective of the new funding methodology was to recognize differences in patient severity, the decision was made to use refined diagnosis-related groups (RDRG) as a transitional severity-of-illness system until a more clinically valid system could be implemented. The RDRG system was developed at Yale University and assigns cases in diagnosis-related groups to one of four severity levels (resource group numbers, RGNs) based on the presence of complications or comorbidities. The ultimate goal of Alberta Health is to develop a prospective, case-based funding system, using standard costing methodologies across the province. To accomplish this, the province needs to develop an Alberta database of comparable costs, by meaningful patient groups, based on actual experience. While the targeted implementation of the guidelines in hospitals across the province will provide the necessary case costing information, this will take a number of years. In the interim, relative resource weights were developed for each RGN, using service intensity weights (SIWs) as proxies for case costs. (SIWs were developed by the Hospital Medical Records Institute, are based on 1985 New York State charge data, and represent the relative average cost per day for each case mix group, CMG.)

Relative resource weights are calculated by multiplying the inlier average length of stay by the SIW for each case within the group. For example, an RGN with a weight of 2.00 is assumed to consume twice as many resources as one with a weight of 1.00.

Hospital Performance Index

The methodology used by Alberta Health each year to derive inpatient funding adjustments for acute care hospitals greater than 60 beds, is briefly described in figure 1, right.

* Predicted Cost/Case

A predicted cost/case is calculated for each hospital by summing the predicted cost/case for inliers and outliers. Inlier predicted cost/case is calculated for each hospital by multiplying the number of cases in each RGN for the year by the RGN weight (derived above) and dividing the sum by the total number of cases. Outlier predicted cost/case is calculated by identifying the number of days over the trim point for each RGN, multiplying by the SIW and a discount factor, and dividing by the total number of cases. The discount factor is .3 for RGNS with a severity of `0' and `1', and .7 for RGNS with a higher severity (i.e., `2' or `3').

The term predicted cost/case is somewhat misleading, as relative weights rather than realistic costs were used in the calculation costs. (The only way to convert relative weights to a meaningful dollar value is to calculate an average cost per relative weight.

* Overall Predicted Cost/Case (Adjusted)

Each hospital's overall predicted cost/case is determined by multiplying its predicted cost/case by an adjustment factor. Alberta Health determines an annual adjustment factor for each hospital, based on hospital size and the number of medical interns and residents, as studies have indicated that these factors affect cost. More recently, the hospital size variable was replaced by the number of adjacent diagnosis-related groups (ADRGs), which is believed to reflect the range of services provided by each hospital. Hospitals with a greater scope of services, and more interns and residents per bed, receive a higher adjustment factor.

* Average Cost/Case

Hospital inpatient operating expense is divided by the total number of cases used in the predicted cost/case calculations to estimate an actual average cost/case. Because of the current limitations in hospital cost accounting, some method was needed to separate the inpatient proportion of hospital operating expense. Alberta Health uses a modified Evans-Barer model. Essentially, a series of formulas determine the proportion of inpatient and outpatient expense for known functional centers (e.g., nursing inpatient, ambulatory care, and diagnostic and therapeutic services), which are then applied to the functional centers where the distribution is unknown or not applicable (e.g. administration and support services).

* Hospital Performance Measure (HPM)

A hospital performance measure (HPM) is calculated for each hospital by dividing its overall predicted cost/case by its actual average cost/case.

* Hospital Performance Index (HPI)

Performance relative to the provincial mean is reflected in the calculation of the hospital performance index (HPI). The HPM of each hospital is divided by the mean performance measure of all hospitals and multiplied by 100. An HPI of 100 indicates that the relationship between overall predicted cost/case and actual cost/case is at the provincial average. An index above 100 suggests that the hospital has performed better, or more efficiently, while one below 100 suggests that the hospital used more resources to treat its patients, in relation to the other hospitals in the province.

The Funding Adjustment

The funding adjustment for the upcoming year is determined by the degree to which the hospital's HPI is above or below the mean. For example, if a hospital's estimated inpatient proportion of funding is $25 million and its HPI score is 105, funding would be adjusted upward by 5 percent, an additional $1.25 million. Similarly, if the hospital's HPI is 96, funding would be reduced by 4 percent, $1 million. During the first few transitional years of the system, these funding adjustments were restricted to a certain percentage of the hospitals' total funding.

Understanding the Variables

The variables under the control of the hospital include:

* Actual operating costs, including provision of programs and services accounting for these costs.

* Case mix group or, more specifically, the distribution of cases by RGN.

* Length of stay of cases within each RGN cell.

Some of the actions taken by hospitals to raise their HPI scores include:

* Reducing overall operating costs through a variety of strategies, including corporate mergers with other hospitals, rationalizing and integrating services, closing beds, contracting out support services, and general downsizing.

* Maximizing the severity rating of cases through improved identification of comorbidities and complications, and greater specificity of discharge diagnoses.

* Reducing the number of cases with lower RGN weights, through greater use of day surgery and other ambulatory care.

* Shortening lengths of stay through preadmission testing and earlier discharge practices, thereby reducing actual operating costs through the elimination of the marginal costs of the reduced days.

The Utilization Management Challenge

Managing a hospital on the basis of a single index number, the Hospital Performance Index, which is calculated annually, and on data that are more than two years old is one of the most challenging hurdles facing Alberta hospitals today. These hospitals need better management information to get a handle on the impact of specific case mix groups and physician practices on the overall HPI and, as a result, on the hospital's funding.

First, many hospitals have addressed the issue of timeliness by calculating their own HPI scores on an ongoing or more current basis. This is not a difficult process, but it does require the assumption that RGN case weights and provincial average HPM have remained unchanged. The resulting HPI models are an approximation rather than a definitive performance index. Figure 2, below, illustrates the results of Wetaskiwin Hospital's modeling efforts.

Second, hospitals have more closely monitored length of stay outliers to identify case mix groups or physicians with a significant negative impact on the hospital's HPI. For example, figure 3, page 32, shows the average number of days over the LOS trim by physician, over time, and clearly highlights those situations that should be followed up in more detail. Alternatively, LOS monitoring may focus on case mix groups, such as the number of cases and the number of days over the trim point, by physician and by RGN.

Third, hospitals have been taking steps to change their case mixes. An analysis of changes in the number and percentage distribution of cases is useful to monitor the effects of operational or clinical changes on case mix.

Under the current inpatient funding scheme, possibly the best information to strive for is the calculation of an HPI equivalent, by physician and by case mix group. The formula for this is quite simple. If actual case costing information is available, each RGN weight could be multiplied by the hospital's adjustment factor and divided by the actual average cost per case for those specific cases. The result may then divided by the most recent provincial mean RPM to approximate an HPI equivalent for each group of cases.

In order to manage in this complex funding environment, Wetaskiwin General Hospital has taken advantage of its ability to cost services provided to individual patients and has developed a number of new management reporting and analysis capabilities, including the calculation of HPI scores by physician and RGN cell. Figure 4, left, and 5, page 33, summarize the HPI performance of the hospital's high-volume case mix groups for all physicians and the HPI performance of individual physicians for a selected high-volume case mix group. Whether produced by physician or by case mix group, such reports readily highlight areas to be investigated in more detail. The goal is to present information to physicians in a useful manner, so they can respond with changes in practice behavior.

In addition to assessing relative efficiency and managing resources, resource utilization information is also used to assess the role of the hospital and to evaluate new or existing services. Cost comparisons with other facilities are particularly useful for this purpose. Financial information on the costs of providing services must be presented, however, in a manner that is meaningful to both management and clinicians. Figure 6, for example, compares the actual costs of treating pediatric patients in Wetaskiwin Hospital with those of a large tertiary teaching facility for the same RGNS.

Summary

Two future activities are required at the Wetaskiwin Hospital to complete the utilization management process. The first is to project the types and volumes of cases that should be treated and their anticipated severity of illness, using historical information and strategic planning initiatives. The second is to develop clinical profiles or standard expectations of resource use and outcomes.

With these in hand, Wetaskiwin Hospital will be in a better position to monitor whether it is serving the most appropriate types of patients and whether available resources are being used efficiently and effectively toward the desired outcomes.

References

[1.] Jackson J. "Setting the Stage: A Framework for Utilization Management." In Utilization Management: A Tool for Informed Decision Making. Proceedings of the National Conference on Utilization Management; Oct. 14-16, 1990. Ottawa, Ontario, Canada: Canadian Medical Association, 1992, pp. 9-13. [2.] The MIS Group. Guidelines for Management Information Systems in Canadian Health Care Facilities. Ottawa, Ontario, Canada: The MIS Group, 1985, 1992 (Rev.).

Premkumar Chengalath is Chief of Staff, Wetaskiwin Health Care Centre, Wetaskiwin, Alberta, Canada, Kim e. Walker, MHSA, CHE, is a Principal with the Arlington Consulting Group, Edmonton, Alberta.