Although the goal of providing care is the same regardless of the payment arrangement, often the payment model requires different approaches to how that care is both provided and particularly how health services are documented. The underlying payment rationale of managed care and Medicare Advantage plans is providers are reimbursed for value and outcomes rather than the volume of services. The reimbursement model attempts to take into account the cost differences between treating different types of patients and adjusting the reimbursement through risk factors such as age, gender, chronic conditions as well as overall health based upon diagnosis codes to reach a risk score for each patient which translate into payments for that patient’s care. The entire model is designed to prevent plans and providers avoiding higher risk patients by compensating plans and providers more to cover the cost differences.
The actuarial categories and formulas used to arrive at a risk score aren’t perfect and there are gaps in the information used to estimate the costs and therefore the weight of certain risk factors. Some of those gaps may be intentional due to fraud concerns. A recent article in the Journal of The American Medical Association makes the argument that the model does not sufficiently account for some diagnoses such as depression and dementia which are calculable.
The argument made by CMS against relying on those diagnoses for calculating the risk score is the potential for fraud; that those factors are not as objective as other diagnoses and are therefore subject to “gaming” of risk scores by providers to increase reimbursement. The argument is not that those diagnoses don’t affect the cost of treatment, but that the CMS does not want to make it easier for providers to cheat to get reimbursement. The gaming of risk adjustment has become a substantial issue for many providers and insurers because of the amounts at stake.