Profiting during chaos.

Tools for improving financial performance under PPS 

IT NOW APPEARS THAT THE SKILLED NURSING facility PPS will continue to be modified continuously as HGFA tries to improve the accuracy of the SNF payment system.

The types of changes that HCFA continues to propose are major modifications to the system, not minor "adjustments", as is often the case in PPS changes. And, these modifications, although intended to improve the payment system by providing more accurate reimbursement for certain types of SNF patients, may not result in improved financial results for facilities because of limitations in the capability of the payment system to recognize variations in patient costs.

Unfortunately, the result of the PPS modifications may be to introduce further chaos into the payment system, making it even more difficult for operators to understand and use. Even more important, however, is the fact that as the system's complexity increases, its reliability as a tool for providing reimbursement that is related to the costs of treatment for patients does not necessarily improve.

The major reason for this problem is that the variations in severity of medically complex patients admitted to nursing homes are not well characterized by the Resource Utilization Groups (RUGs) system, and this leads to use of average reimbursement amounts, which are not clearly related to the costs of care. While the "new" version(s) of the PPS will provide more reimbursement for medically complex patients, which is an improvement, this does not assure that such reimbursement will in fact cover patient costs.

Can SNF financial performance be improved?

Is it possible to improve the financial performance of SNFs despite the chaotic nature of the payment system? The answer to this question is a qualified yes, if SNFs are willing to adopt techniques that can adjust for the limitations of the PPS system. They would have to forecast patient costs in advance and case-manage patients so margins are maintained based on PPS reimbursement amounts. This adjustment requires that one major principle be understood and accepted: that the SNF PPS system, based on the RUGs methodology, is not a sound basis for estimating the costs of SNF patients. Once this principle is accepted, SNFs can proceed to adopt tools that will more correctly estimate patient costs and provide information to assure improved profitability.

Fundamentals of cost forecasting

Forecasting is useful when financial predictions are needed and where multiple complex factors influence the predictions. Forecasting methods are particularly applicable to health care because multiple patient conditions (factors) directly influence the amount and type of treatment patients receive; this determines treatment costs. If the most predictive patient factors can be identified and statistically evaluated, reasonable predictions of costs and clinical outcomes can often be made.

Forecasting methods are particularly well suited for pre-admission and admission predictions but can be used for predicting costs as patient treatment takes place. Cost-forecasting is particularly important in situations where reimbursement is fixed or capitated, but treatment costs vary based on factors such as patient diagnosis, co-morbidities, and severity. Often, in SNFs and other types of post-acute care, PPS systems are not designed to evaluate patient medical conditions correctly in relation to required treatment and costs. This leads to the potential of less-than-adequate reimbursement for certain patients. in such circumstances, early knowledge of expected patient costs can enable appropriate patient management to assure proper treatment within both cost and reimbursement limits.

Overall, because of the relatively high variation in patient costs, cost forecasting is one of the few ways that a SNF can both protect itself against unanticipated and uncompensated expense and improve efficiency of patient management to assure margins.

Cost-forecasting consists of three basic functions: (1) classifying patients; (2) developing statistical prediction formulas; and (3) producing cost, margin, and outcome predictions.

Classifying patients involves using certain, relatively limited, data on patient demographics, diagnoses, and functional level and data on the costs of treatment.

Prediction formulas involve using standard statistical procedures to analyze clinical and cost data.

Producing predictions consists of generating data on expected cost per day and per episode of care, expected reimbursement based on RUG group(s) or managed care payment rates, anticipated margins (P/L), and, if desired, clinical outcomes.

Cost-forecasting methods for SNFs produce results that are generally as accurate as the acute care hospital DRG system. Even better results are available for many SNF rehabilitation patients. One of the great advantages of cost-forecasting is that it can be programmed into software. This means that patient cost predictions and margin analysis can be produced virtually instantaneously, including using portable computing devices.

An additional advantage is the capability to produce continuously updated management reports showing the financial results of a changing case-mix over quarterly, semi-annual, and annual periods. Therefore, cost-forecasting provides very early prediction of SNF patient costs and margins and enables management of costly patients (including outliers), so that margins can be maintained. Cost-forecasting provides:

* basic patient diagnosis and severity classification as well as RUG classification

* individual patient level cost forecasts

* patient level reimbursement data

* costs by reimbursement group

* cost predictions based on facility cost data

* P&L by patient type

* multiple venue costing capability

* automated management reports

* internet capability.

The way cost-forecasting works

Cost-forecasting provides financial predictions based on provider- or facility-specific financial and clinical data. Limited patient demographic and clinical assessment data, collected at pre-admission or admission, are input into the system. The output are predictions of expected costs and margins based on a specified patient reimbursement amount based on RUG group or managed care payment rate. Cost-forecasting is available for SNFs and can be developed for medical rehabilitation, long term acute care hospitals, and home health providers.

The data input requirements are minimal and include patient age and sex, and a small number of clinical assessment variables such as diagnosis, an ADL score, and the number of co-existing medical conditions presented at pre-admission or admission. These data elements are included in the Minimum Data Set 2.0 instrument. Generally, keying in the required data requires less than five minutes. Cost data are pre-entered in the system based on facility-specific costs, therefore there is no keying of cost data required.

The most important component of accurate cost-forecasting is a set of statistical prediction formulas, which are also pre-programmed in the system. These formulas are based on research involving thousands of SNF cases from facilities throughout the country. The formulas generate the predictions but the predictions depend entirely on specific patient characteristics and facility costs. The formulas consist of statistical procedures that produce the best possible predictions of costs based on patient and facility data. (See "Making financial predictions," above.)

Uses and results of cost forecasting

Cost-forecasting has multiple uses, including pre-admission patient evaluation, admission patient evaluation, case management targeting, case-mix management, and treatment cost review

The use of cost-forecasting results in maintaining and improving margins, by identifying the costs of treatment compared with reimbursement, prior to or at patient admission, thereby encouraging use of efficient treatment protocols or case management of patients with high resource use. When available, outcomes data can be included in forecasting predictions, thus assuring cost-effective treatment.

Carefully forecasting and monitoring treatment cost leads to improved financial performance under PPS. The great advantage of cost-forecasting is that its availability through software makes it possible for both admissions and case management staff to understand immediately the predicted expense of patients both before and at admission. It also makes it possible to correctly evaluate margins and to manage patient treatment actively to both maintain financial viability under PPS and managed care payment systems as well as assure quality outcomes.

Malcolm H. Morrison, PhD, is president and CEO of Morrison Informatics Inc., a Mechanicsburg, Pa.-based information technology and data analysis consulting firm specializing in post-acute care.

Making financial predictions

The prediction consists of the following information, which is produced immediately after entering the few required patient data elements:

* cost per day during treatment (specific cost components can also be provided)

* reimbursement per day (determined by RUG group or managed care rate)

* length of stay per episode of care

* total cost per episode of care

* total reimbursement per episode of care

* margin per day and per episode.

The system also provides management reports, specifically including:

* cost by RUG report

* cost by RUG and diagnosis report

* cost by severity report

* detailed cost components by RUG report (pharmacy, laboratory, therapies)

* detailed cost components by RUG and diagnosis report

* detailed cost components by severity report

* customized facility predicted vs. actual per diem and per episode cost by RUG report.