Risk Theory 1

Professional Caregiver Insurance Risk and Average Cost Based Reimbursement Plans: Implications for Nursing

Executive Summary

Thomas Cox RN, MS, MSW, MSN

Doctoral Candidate

School of Nursing

Virginia Commonwealth University

1220 West Broad Street

Richmond, VA 23231

"In today's rapidly changing medical marketplace, managed care plans are not the only entities assuming risk for the care of enrollees through capitation. Increasingly, managed care plans are transferring this risk to their primary care and specialty physicians by paying them on a fully or partially capitated basis. Although capitation provides a strong incentive for physicians to provide cost-effective care, there are concerns that capitation may place some physicians at considerable financial risk." (Anderson & Weller, 1999)

Introduction

Capitation contracts (CCs) and average-cost based reimbursement Plans (ACBRPs), such as Diagnosis Related Groups (DRGs) have existed for decades. They create risks for health providers and consumers that are not incidental but integral, though analysts have incorrectly identified and appraised these risks. These financing mechanisms have dire effects on providers and consumers, and have also affected the insurance industry, its operations and regulation. The long-term consequences, however, have been little explored and this paper will address some of these issues from the standpoint of nursing and nurses. Viewing ACBR plans as insurance contracts, reveals serious flaws. Such a view will help to demystify these mechanisms, open new vistas of research activity, and potentially contribute to ending the use of these harmful efforts at cost containment. The unequal knowledge bases of the contractees also raise very serious ethical, legal, professional, and business concerns.

Providers engaged in CCs or ACBR plans agree, often involuntarily, to care for clients, trading fixed payments for uncertain costs, creating and assuming many risks; some are predictable, and some are far from predictable. Predictable, though rarely predicted, risks reflect variations in health costs related to factors such as age, gender, employment status, or chronic illness status. Other poorly analyzed risks include the operating characteristics of the provider, which vary from unit to unit, shift to shift, based on patient acuity, staffing, resource management, and availability of needed supplies. Other risks include: self-selection biases, financial adversities when borrowing, and liabilities due to unplanned resource needs. Small providers tend to poorly manage these risks. Few discussions however, direct attention to the major risk source, that inadequately resourced providers have entered the insurance business and adequately resourced insurers left the insurance business with no regulatory oversight or action.

The benefit of insurance is the accumulation of risk so that the Law of Large Numbers (LLN) and the Central Limit Theorem (CLT) make the aggregate risk highly predictable, benefiting both insurers and insureds. Role reversal, as occurs in ACBRPs damages legitimate insurers, providers, consumers, and the health care system through: business failures, consumer dissatisfaction, provider consolidations and takeovers, reduced access to care, and delayed and deferred patient diagnosis and treatment. Entities that use ACBRPs to avoid the benefits of risk consolidation include: insurers, managed care organizations (MCOs), HMOs, unions, and governmental entities. Professional caregivers are incapable of performing the conflicting roles of insurers, gatekeepers, and service providers that their entry into the insurance business entails. Insurers and government render valuable service to society when they aggregate risk and reduce uncertainty for themselves and their insureds. ACBRPs fail to provide such a service. Increased risk to, and risk assumption by providers has radically altered the healthcare system, undermining professional values, eroding consumer confidence, and resulting in provider consolidations to the detriment of the nation's confidence, health, well being, and preparation for disasters and other unexpected demands on the health care system.

Insurance and Probability Issues

Insurers manage risk better than insureds by being adequately capitalized to meet normal fluctuations from period to period and well enough capitalized to handle unusual fluctuations in costs from period to period. Insurers also benefit from high volumes of business, greater geographic risk spread, and multi-line insurance diversification. The CLT and the LLN are the guiding principles in insurance operations because insurers predict costs more accurately by writing large numbers of identical policies.

This benefit from the CLT and the LLN is the same as the statistical sampling effect of taking ever-larger samples from an unknown distribution. As the sample size increases, the variability in average values decreases. Increasing the number of sampling units reduces the range of error in estimating the average costs for all the units. This is also the principle behind the use of the normal distribution. In terms of portfolios of clients, smaller organizations, such as community hospitals, private practitioners, nursing homes or extended care facilities that contract to provide services to relatively small numbers of clients have higher variability in their results than the insurers who transfer the risk these clients present. This greater variability is predictable with a few, conservative assumptions. These more variable outcomes are due solely to the differences in the number of clients in the sub-portfolio compared to the full portfolio in reverse of the usual insurance relationship. This increases the probability of both good and adverse financial outcomes for the smaller providers who receive disaggregated collections of insured clients. This effect is also independent of any discrepancies in the average values of the particular portfolio transferred. Clearly, if the members of the sub-portfolio are sicker than expected the provider has automatically been exposed to higher costs than those they were paid to accept. Insurers can manage such variability because they are obligated to maintain assets that are far greater than the losses they are intended to cover, but providers: physicians, hospitals, rehabilitation centers, laboratories are neither obligated to maintain such reserves nor do they have the assets available to them to meet such requirements.

The normal distribution is a good model for statistical issues in a wide variety of situations (Robbins & Van Ryzin, 1975). However, the normal distribution is a poor model for insurance because of the small numbers of very large claims and the large numbers of very small or zero claims that characterize insurance contracts. Nevertheless, the explanation in this paper uses the normal distribution, a conservative model, to clarify and simplify issues regarding variability and risk. The normal is conservative in two ways: the risk of better than average experience equals the risk of worse than average experience; and the normal has smaller sampling variances than other, more appropriate distributions. The risks of adverse financial experience are higher when modeled with other distributions. The approach taken will be a comparison of the likelihood of events distant from the expected (mean) value in provider groups with the probability of equally distant, from average, experience for an insurer. The different spread around the average due solely to the disparity in portfolio sizes between providers and insurers is the focal point of this paper.

Assume that there is an amount of the insurer's premium that is the actual cost of providing care to the insureds through providers (i.e. provider cost) and that it is distributed as N(.70, .05) (normal distribution with a mean loss ratio of 0.70 and a standard error of 0.05). The assumptions about parameter values are not important here. The rest of the premium covers profits, expenses, loss adjustment services, and other necessary costs. If the insurer retains risk, the probability that the insurer will experience a loss ratio other than 0.70 is easy to determine, the probability that the insurer's losses in any year will exceed 0.75 is 0.1587. If the insurer's premium incorporates a 5% profit margin, the insurer will be unprofitable less than 16% of the time. Now assume that a provider group accepts the average loss cost as payments in a capitation contract. What is the probability that the provider will have a loss ratio in excess of their average payment 0.70? Also, note that consistent with the above assumptions about premium allocation, this represents 100% of their guaranteed income to cover services since the insurer cannot pay out more than this amount. The very first critical observation is that the provider’s experience is as likely to fall above the average as below, there is a 50% chance that the provider will have costs that exceed the amount budgeted, on average, for care delivery.

If the provider accepts 10% of the insurers portfolio, their standard error is .05*(10**.5) or 0.158. The probability that the provider's loss ratio will exceed 0.75 is 0.376. The relative risk of a provider loss exceeding 0.75 is 2.37 times that for the insurer. If the same provider assumes only 5% of an insurer's portfolio the relative risk of a loss greater than 0.75 is 2.59 times that of the insurer due to size differentials alone. Coupled with the provider's limited ability to manage such losses this higher probability of a poor result may be devastating. No extant regulatory bodies determine whether providers can manage such risks. State and federal regulations require insurers to maintain sufficient assets to weather unexpected results. No similar requirements exist for providers that accept risk laden ACBRPs. While increased vulnerability to high losses is offset by a greater probability of achieving a lower than expected loss, the advantages of increased profitability are offset, among providers, by the risk of going bankrupt. Most providers would not be willing to risk bankruptcy just for an equal chance of greater profits. Few providers or consumers realize that their nominal insurers were no longer managing these risks. Both providers and consumers were simply unaware that insurers had actually transferred their insurance role to providers. Providers can only manage ACBRP agreements by targeting services at a lower level then the loss costs in the payments they are given by insurers. The only mechanisms available to providers to accomplish this are by limiting care delivery, threatening the health and well-being of their clients and increasing . Uncompensated risk assumption is not viable over the long-term.

This example demonstrates that the provider groups have far greater probabilities of extreme losses than insurers. In addition, provider groups are also financially weaker. Where well-capitalized insurers can withstand several years of extremely high losses, many small providers cannot survive a single month without precipitating a financial crisis. Another dissimilarity between insurers and providers that works against risk assumption by providers is that insurers benefit by reducing their expenses when handling a large number of relatively similar policies. Writing more policies introduces small increases in operating costs for insurers. Providers expense costs increase when dealing with dissimilar CCs and ACBR plans that increase the complexity and inefficiency of their operations. The way for a capitated provider to manage the risk of adverse financial experience is by targeting average costs to be lower than guaranteed reimbursements. To do anything other than this invites long-term financial ruin. While most providers will have trouble with this and will not willingly travel these paths, poor planning for and poor implementation of these contracts can quickly place providers in financial jeopardy necessitating such measures. Other features of these contracts, such as bonuses for limiting the use of expensive diagnostic and treatment protocols place providers in an ethical and legal conflict where it is difficult, if not impossible, to act in the best interest of patients and act in the best interest of their group.

Relative Risk of Adverse Experience for Providers versus Insurers

Table 1 below details the probabilities of losses greater than or equal to a given loss ratio for the insurer, the provider and finally, the relative risk for the provider compared to the insurer. The table is based on an expected loss ratio of 0.70, a standard error of 0.05 for the insurer and the assumption that the provider group accepts 5% of the insurer's clients. At an actual loss ratio of 0.70 there is no risk differential. However, as the actual loss ratio increases, the fact that the standard error is lower for the insurer, the relative risk becomes greater as the experience moves away from 0.70. The table clearly demonstrates that the provider groups have far greater probabilities of extreme losses than is true for the insurer. This is not all that comes into play however. The provider groups are also weaker financially. Whereas an insurer may be able to withstand several years of extremely high losses, many small providers move from month to month with little margin for higher than expected losses.

Table 1

Relativity Probabilities of Adverse Financial Results

Provider accepts 1/20 of insurers portfolio

Loss / Insurer / Provider / Relative
Ratio / Risk / Risk / Risk
0.70 / 0.5000 / 0.5000 / 1.0000
0.71 / 0.4207 / 0.4822 / 1.1460
0.72 / 0.3446 / 0.4644 / 1.3480
0.73 / 0.2743 / 0.4466 / 1.6290
0.74 / 0.2119 / 0.4290 / 2.0250
0.75 / 0.1587 / 0.4115 / 2.5940
0.76 / 0.1151 / 0.3942 / 3.4260
0.77 / 0.0808 / 0.3771 / 4.6700
0.78 / 0.0548 / 0.3603 / 6.5740
0.79 / 0.0359 / 0.3437 / 9.5650
0.80 / 0.0228 / 0.3274 / 14.3890
0.81 / 0.0139 / 0.3114 / 22.3960
0.82 / 0.0082 / 0.2958 / 36.0780
0.83 / 0.0047 / 0.2805 / 60.1760
0.84 / 0.0026 / 0.2656 / 103.9550
0.85 / 0.0013 / 0.2512 / 186.0540

Insurer-Provider Financial Strength Differentials

ACBR plan contracts are clearly biased in favor of the well-being of the larger organization. State and Federal regulations require insurers to maintain their financial capacity to withstand risk. Insurers hold considerable assets that cover their needs in years in which they suffer unusually high losses. GAAP (Generally Accepted Accounting Principles) and FASB (Financial Accounting Standards Board) accounting standards compel insurers to maintain adequate liquidity that allows them to weather the vicissitudes of the insurance business. Small providers, on the other hand, have no such requirements and, in fact, rarely have this capability.

In truth, many providers enter into these contracts because they are financially vulnerable rather than financially strong. A small provider that is not meeting its business costs, or operating at lower profit margins then possible, may be tempted to enter such a contract believing that they will become profitable. They assume this increased vulnerability to high losses without a real appreciation of the fact that the increased risk of loss may mean the difference between a business failure and a marginally profitable and continuing business. That the smaller providers are subject to this higher dispersion of risk is clear. It is difficult to understand the lack of recognition that these contracts force providers into roles as insurers (Cox, 2001).

Can Providers Manage Risk with Stop Loss Reinsurance?

Some might argue that providers can manage the risks of adverse financial experience by securing reinsurance. This is a flawed view. In a steady state condition, market forces will drive the price of insurance down to a minimum. The insurer transferred the risk to the provider for the average cost. The average cost does not, and cannot provide a risk premium to the provider because the cost of doing so exceeds the loss cost provision in the insurer's premium that the insurer received, because the risk increased when the insurer passed it on. In order to purchase reinsurance, the provider must further reduce their target service delivery provision level, engendering further reduced service, diagnosis and treatment to their patients. Purchasing stop-loss insurance simply will not work. Such a scenario asks the provider to provide insurance coverage abdicated by the insurer and then to pay a reinsurer to cover the risk the first insurer was to aggregate and maintain.

Legal Issues Regarding Contracts of Adhesion

Without intending to enter into legal discussions or ethical discussions, it is clear that these contracts exist between parties that have unequal understandings of the risk theoretic consequences of these contracts. Many private practices, hospitals, and nursing homes have become financially vulnerable because of these inherently unfair financial contracts. In many cases, these provider organizations have been faced with "Take it or leave it" contracts imposed by insurer organizations. These contracts have negatively affected providers, legitimate risk assuming and retaining insurers and the public. It would appear appropriate for litigation to test the validity and fairness of these contracts in the courts and, if deemed appropriate, that victims of these contracts, providers and disenfranchised consumers, be compensated for their losses.

Conclusion

Average-cost based reimbursement plans are similar to insurance contracts in terms of the risk transfer. They are dissimilar to insurance contracts in that the party accepting the risk for the average cost is less capable of managing the risk. Provider contractees are smaller, more financially vulnerable and harmed by the greater probability of excessive losses they face. Using a normal distribution as an approximation to the experience under a CC, this author compared risk susceptibility between providers and insurers. Capitated health care providers face higher probabilities of financial loss and this can only be moderated by the delivery of a lower level of service than paid for in these agreements. Over time, one would expect these contracts to result in necessary reductions in both the quantity and quality of services. Properly viewed as reinsurance agreements rather than service contracts, ACBR plans will result in financial ruin, takeovers and consolidation of health providers as well as reductions in available services, the effects observed in the past two decades.