Denying payments for never events is based on the illusion that these events are always avoidable. A better system is to use health care "warranties."
 |
 |
| Donald E. Fry, M.D. | Michael Pine, M.D. |
Neverland. It conjures up thoughts of Walt Disney, Peter Pan, Tinkerbell and maybe Michael Jackson. Neverland, as created by Scottish author J.M. Barrie, is the world where children never grow up, a metaphor for escape from reality.
Now, Neverland has invaded the real world of health care reimbursement. The Deficit Reduction Act of 2005 directed the secretary of Health & Human Services to phase out Medicare payments for medical complications that result from unacceptable clinical practices. Ten sets of complications have been designated as never events by the Centers for Medicare & Medicaid Services (CMS). Effective Oct. 1, 2008, hospitals no longer are reimbursed for the care of these never events. CMS has indicated that it intends to expand its list, and commercial insurers have stated that they too will deny payment for the care of never events.
Never Events Happen
Some complications such as wrong-site surgery and retained surgical sponges can appropriately be classified as never events; but never events such as post-operative infections, venous thrombosis after total joint replacement and pressure ulcers may occur even when patients receive the best care available. Furthermore, published medical literature has shown that sicker, more complex patients are more likely to experience such complications.
Given these realities, what will be the consequences of refusing to pay for the care of a broad array of never events? Kerry Weems, the previous administrator of CMS, contends that this new policy will reduce the number of hospital complications and the cost of hospital care. This projection is based on the unrealistic assumption that there will be no unintended adverse consequences that arise from the denial of payment for the care of improperly classified never events.
If payment is denied for the care of sometimes unavoidable complications, these costs must be shifted to other billable services for hospitals to remain solvent. To limit their financial liability, hospitals may deny care to identifiably high-risk patients. Safety net facilities that already treat higher-risk patients will further be economically disadvantaged. Designation of these understood and expected complications of care as being never events has profound liability implications for hospitals and physicians. These are economic realities that can be ignored only in Neverland.
The Health Care Warranty
An equitable approach that recognizes the limitations of modern medical care would be for hospitals to offer and provide CMS and other payers with a health care "warranty." The warranty is a well-established business practice that is immediately applicable to hospital care. The warranty payment would be added to the payment for the care of uncomplicated cases and should be calculated separately for each patient to ensure that the warranty costs are accountable for variations in patient severity. The costs of a warranty for an individual patient can be derived prior to treatment by computing the risk-adjusted rate at which covered complications occur in high-quality hospitals and the costs associated with these complications.
To create a warranty, we select a reference group of hospitals that have appropriate standards of quality coding, representative rates of complications, and cost profiles that are within national standards. An aggregate or composite measure of outcome is essential to determine quality and cost of outcomes objectively. We have defined the adverse outcome as death or a prolonged risk-adjusted length of stay for inpatient care. This combination provides an excellent objective and aggregated surrogate for serious hospital-acquired complications.
Our recent research has demonstrated a strong correlation between measured adverse outcome rates and excess costs of care, because complications that do not extend lengths of stay generally are not severe enough to affect the cost of care materially. (See Fry DE, Pine M, Jones BL, Meimban RJ, "Adverse Outcomes in Surgery: Redefinition of Post-operative Complications," in Am J Surg 2009; 197:479-84.)
Valid comparisons of adverse outcome rates to identify hospitals with good outcomes can be performed with readily available claims data or claims data enhanced with a few selected clinical data elements. (See Pine M, Jordan HS, Elixhauser A, Fry DE, Hoaglin DC, Jones B, Meimban R, Warner D, Gonzales J, "Enhancement of claims data to improve risk adjustment of hospital mortality," in JAMA 2007; 297:71-76. See also Fry DE, Pine M, Jordan HS, Elixhauser A, Hoaglin DC, Jones B, Warner D, Meimban R, "Combining administrative and clinical data to stratify surgical risk," in Ann Surg 2007; 246:875-85.)
Predictive models for rates of adverse outcomes and their associated costs are then derived from data of the reference group of hospitals. Total costs of care for each case are computed using billed charges and cost-to-charge ratios; more precise data from internal accounting systems (e.g., charge masters) can also be used. Predictive models of cost are developed using cases without adverse outcomes. Excess costs of the adverse outcome are determined as those costs that exceed the uncomplicated care of a patient with the same risk profile.
The warranty for an individual patient can be computed as the product of the probability the patient will have an adverse outcome and the predicted cost of the adverse outcome. Stop-loss provisions may be incorporated into warranties to limit hospitals’ financial exposure when catastrophic adverse outcomes occur.
Hospitals and physicians that have better-than-predicted results will benefit financially from the warranty. Suboptimal performance, including the occurrence of true never events, will result in costs that exceed the warranty payment. Quality improvement and good economic stewardship will be rewarded without unrealistically shifting costs for unavoidable adverse outcomes.
The risk-adjustment models used in the warranty calculation should be revisited periodically to reflect changes in medical technology, clinical practices and available data. One important enhancement as better post-discharge data becomes available would be to incorporate readmissions and post-discharge ambulatory interventions (e.g., drainage of wound abscess) into the adverse outcome measure. To ensure the integrity and accuracy of any payment system that incorporates warranties, predictive modeling and subsequent analyses should be performed by independent organizations with experience in sophisticated measurement of risk-adjusted outcomes.
Time to Get Real
Everyone must accept responsibility and accountability for the costs of complications of patient care. Hospital leaders and physicians must be more accountable for the quality and efficiency of delivered inpatient health care. Patients and payers need to recognize that there are unavoidable and negative consequences to modern medical care. It is proper for CMS and private insurers to influence payments for suboptimal rates of complications or for those events that are truly unacceptable. However, it is time for health care payers to leave the unrealistic world of Neverland, and embrace payment strategies grounded in clinical reality rather than wishful thinking.
Donald Fry, M.D., is the executive vice president of Michael Pine and Associates in Chicago and adjunct professor of surgery at Northwestern University, Evanston and Chicago, Ill. Michael Pine, M.D., is the president of Michael Pine and Associates and a lecturer in the Department of Medicine at the University of Chicago.
This article 1st appeared on August 17, 2009 in HHN Magazine online site.
To respond to this article, please click here.
