|
MARKETWATCHMedicare Payment For Selected Adverse Events: Building The Business Case For Investing In Patient Safety
This study estimates that Medicare extra payments under the hospital prospective payment system (PPS) range from about $700 per case of decubitus ulcer to $9,000 per case of postoperative sepsis in the five types of adverse events identifiable in Medicare claims. Medicare extra payment for the five types of events totals more than $300 million per year, accounting for 0.27 percent of annual Medicare hospital spending. But these extra payments cover less than a third of the extra costs incurred by hospitals in treating these adverse events. We conclude that both Medicare and hospitals gain financially by improving patient safety.
BESIDES HARMING PATIENTS, adverse medical care events have major financial consequences.1 But who bears the extra costs for treating adverse events is not well understood. The general perception is that providers are reimbursed for treating adverse events that occur when patients are under their care, shifting the burden to patients, employers, health plans, Medicare, Medicaid, and other payers. The evidence is unequivocal that hospitals charge more when an adverse event occurs during hospitalization.2 However, in a system with multiple payment mechanisms and complicated provider-payer contractual arrangements and price concession schemes, it is not at all clear who actually covers the extra costs. Understanding the financial aspects of adverse events, especially the financial incentives embedded in today’s health care system, is critical to the proper design of programs and financial incentives aimed at improving patient safety. This study assesses how much Medicare under its diagnosis-related group (DRG)–based prospective payment system (PPS) pays hospitals for five selected types of adverse events identifiable in Medicare claims, and how much of the extra costs are absorbed by hospitals.3 The objective is to provide the first insights into the intricate financial relationships surrounding adverse events and illustrate the business cases for both Medicare and hospitals to invest in patient safety.
Data and variables. Our primary source of data was the 2002 Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS), developed by the Agency for Healthcare Research and Quality (AHRQ). The 2002 NIS file contains about 7.85 million uniform hospital discharge abstracts from 995 short-term, general hospitals in thirty-five states, approximating a 20 percent stratified sample of nonfederal acute care U.S. hospitals.4 The NIS includes variables on sources and types of admissions, DRGs, fifteen diagnosis codes from the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM), fifteen ICD-9-CM procedure codes, procedure days from admission, discharge status, length-of-stay, total charges, patients’ demographic characteristics, insurance coverage, and a few hospital-characteristic variables. The database also includes sampling weights and design variables for generating national estimates. Calculating Medicare payment under the hospital PPS required three data files from the Centers for Medicare and Medicaid Services (CMS): (1) DRG relative weights; (2) the Area Wage Index and a number of other variables representing variation in cost of living across the country, listed by metropolitan statistical area (MSA) codes, including a single code assigned for rural areas in each state; and (3) the Medicare IMPACT data, which include key variables for payment adjustments for individual hospitals, such as the resident-to-average-daily-census ratio, disproportionate share of low-income patients, urban/rural designation, number of beds, and cost-to-charge ratios.5 The IMPACT file was linked to NIS 2002 by a hospital identifier crosswalk created by AHRQ. Identifying adverse events during hospitalizations. We used the AHRQ Patient Safety Indicators (PSIs) to identify adverse events during hospitalization.6 The AHRQ PSIs include twenty indicators with reasonable validity, specificity, and potential for fostering quality improvement. Inclusion and exclusion criteria were developed to screen out adverse events that are not preventable. Although the validity of these indicators is not fully established, both empirical data and an expert clinical panel agreed that these indicators have sound clinical validity in identifying preventable adverse events.7 Our analysis included the following five PSIs with relatively larger numbers of cases and representing diverse safety concerns. Decubitus ulcer (DU). This includes a DU diagnosis in any secondary diagnosis field and in hospitalizations with stays longer than four days, excluding patients with principal diagnosis of DU; patients with any diagnosis of hemiplegia, paraplegia, or quadriplegia; and patients admitted from long-term care facilities. Iatrogenic pneumothorax (IP). This includes an IP diagnosis in any secondary diagnosis field, except in patients with principal diagnosis of IP, trauma patients, and patients with thoracic surgery or lung or pleural biopsy or assigned to cardiac surgery DRGs. Postoperative hematoma or hemorrhage (H/H). This includes an H/H requiring a procedure after an operating room (OR) procedure, except in patients with principal diagnosis of H/H, patients with the only OR procedure that is postoperative control of hemorrhage or drainage of hematoma, and patients in whom a procedure for postoperative control of hemorrhage or drainage of hematoma occurs before the first OR procedure. Postoperative pulmonary embolism or deep vein thrombosis (PE/DVT). This includes a PE/DVT in any secondary diagnosis field after an OR procedure, except in patients with principal diagnosis of PE/DVT, patients in whom a procedure for interruption of vena cava is the only OR procedure, and patients in whom a procedure for interruption of vena cava occurs before or on the same day as the first OR procedure. Postoperative sepsis (PS). This includes a sepsis in any secondary diagnosis field in elective surgical discharges with length-of-stay of four days or more and an OR procedure, except in patients with principal diagnosis of sepsis or infection and any diagnosis of immunocompromised state or cancer. Calculating Medicare payment for hospital claims. The Medicare hospital PPS pays for inpatient services based primarily on DRGs. The DRG system sorts patients into more than 500 groups, each group of cases having related clinical problems and expected to have similar costs. Each DRG has a relative weight reflecting the relatively costliness of the DRG compared with other DRGs. Medicare PPS payment for inpatient services has five components, base operating payment, base capital payment, an indirect medical education allowance, an adjustment compensating hospitals for treating disproportionate shares of low-income beneficiaries, and an adjustment for cases with unusually high costs relative to the payment rate for the DRG. Formulas and parameters for payment determination are published at the beginning of every federal fiscal year. For this study we used specific formulas from the Medicare Payment Advisory Commission’s (MedPAC’s) March 2002 report to Congress on Medicare payment policy, supplemented by additional details from other sources.8 Calculating Medicare extra payment for adverse events. We first screened the selected types of adverse events based on ICD-9-CM codes recorded in Medicare claims. The DRG for each claim with an adverse event was assigned using a DRG grouper from 3M Inc., and Medicare payment for the claim was calculated.9 In the second step, the ICD-9-CM codes indicating adverse events were removed from the claim, the DRG was reassigned, and payment was recalculated. The second step created a simulated claim mirroring the original claim without the adverse events. This step required a total charge for the simulated claim to determine cost outlier adjustment along with the cost-to-charge ratio obtained in the Medicare IMPACT data. Using the 2000 NIS, Chunliu Zhan and Marlene Miller estimated excess charge attributable to each PSI, including the five types of events studied.10 The total charge for the simulated claim thus was determined by deducting the excess charges from the claimed total charge and inflating to 2002 dollars: simulated total charge equals the claimed total charge minus [1.101 x (excess charge due to PSI)]. In the third step, we examined the differences in DRG designations and Medicare payment between the actual case and the simulated case, and we estimated the changes in payment components and the national total of Medicare payment for the selected adverse events.
Adverse events during hospitalization. The 2002 NIS contained more than 2.5 million hospital discharge abstracts for patients who were age sixty-five or older and who claimed Medicare as their primary or secondary payer. Exhibit 1  shows the national estimates of the five types of events among these Medicare beneficiaries. The event rates vary from one case of iatrogenic pneumothorax to thirty-four cases of decubitus ulcer per thousand discharges at risk.
Changes in DRGs after adverse events. Exhibit 2 shows that among the discharges with adverse events identified, only a small percentage would have changed DRG designations had the events not occurred during hospitalization. Exhibit 2 also shows that the cases with DRGs changed had fewer comorbid conditions, with lengths-of-stay, charges, and mortality several times lower than for the cases with DRGs unchanged.
An examination of diagnosis codes recorded in discharge abstracts showed that the cases with DRGs changed were those without comorbid conditions and where patients did not suffer other complications during hospitalization. Most patients with adverse events had one or more comorbid conditions or other complications, which explained why removing adverse events did not change the DRG designation for more than 90 percent of the discharges with adverse events.
Medicare payment for adverse events.
The Medicare payment for adverse events under the PPS ranged from $735 per case of decubitus ulcer to $8,881 per case of postoperative sepsis (Exhibit 3
Components of Medicare payments. More than half of the Medicare payment for a claim was to compensate basic operating cost, and about 5 percent each was to cover capital costs, indirect medical education, and allowance for treating disproportionate numbers of lower-income patients (Exhibit 4 ). Payment adjustment to cover unusually high-cost patients accounted for the remaining 40 percent. Although the increased payments attributable to adverse events for the first four components were 2.5 percent at the highest, payment adjustments for unusually high costs increased by 15–48 percent when adverse events occurred. The payment for adverse events accounted for 5–20 percent of the average payment for cases with adverse events, more than 80 percent of which were outlier payment adjustments.
Share of Medicare budget paid for adverse events. This study shows that Medicare pays hospitals a substantial amount for adverse events occurring during hospitalization. The estimated Medicare payment for the five types of adverse events accounted for 0.27 percent of the total Medicare hospital spending of $117 billion in 2002.11 The total Medicare payment for adverse events is undoubtedly much higher, since these five types of adverse events are a small subset of all adverse events, and many of the targeted adverse events are not captured by the claims data. Nevertheless, this finding presents a clear business case for Medicare to invest in hospital patient safety.
Business case for patient safety.
This study also suggests a business case for hospitals with substantial shares of Medicare patients to improve patient safety. Our findings show that hospitals get no additional payment under the Medicare PPS in 48 percent (postoperative sepsis) to 80 percent (decubitus ulcer) of the cases when adverse events occur. In these cases, hospitals have to absorb the extra costs in treating the adverse events. Although it is difficult to estimate how much hospitals absorb the extra costs under the PPS, it can be inferred. Previous studies estimated excess charges per case at $10,845 for decubitus ulcer, $17,312 for iatrogenic pneumothorax, $21,431 for postoperative hematoma or hemorrhage, $21,709 for postoperative pulmonary embolism or deep vein thrombosis, and $57,727 for postoperative sepsis; and inpatient cost-to-charge ratios had a national weighted average of about 45 percent in 2002.12 Based on these estimates and the extra payments presented in Exhibit 3
Implications for the Medicare PPS.
This study also helps inform changes in Medi-care PPS policies. First, the current PPS does not distinguish between comorbid conditions that are present at admission and complications or adverse events that occur during hospitalization in assigning the DRG and in determining payment. Our findings support MedPAC’s 2005 recommendation that the CMS should require hospitals to identify on their claims forms which secondary diagnoses were present on admission.14 This information would allow Medicare to distinguish between payment for treating comorbid conditions, which is reasonable, and payment for treating adverse events, which is questionable. Second, our findings support and inform the CMS’s efforts in developing pay-for-performance (P4P) schemes for patient safety. Our study suggests that some patient safety measures could form the basis of a P4P plan, and a national benchmark could be estimated, as presented in Exhibit 1 Study limitations. Some limitations should be noted. First, we studied only five selected types of adverse events identifiable through Medicare claims. The large volumes of the claims and the ICD-9-CM codes recorded in the claims made our analysis possible; in the meantime, our findings should be interpreted in light of the limitations of the data and the methods.16 In general, missing codes, erroneous codes, and lack of codes for specific safety events are expected. As a result, the reported adverse events represent only a subset of patient safety events and are subject to identification errors. More seriously, despite the best effort to identify only preventable adverse events, some of the adverse events identified in Medicare claims might not be preventable. Nevertheless, this approach has been well established and is the best available method for our purpose. Second, we calculated extra payment for the five selected adverse events made by Medicare under its DRG-based PPS. The finding is not generalizable to other payers or payment systems. As our analysis showed, the extra payments vary greatly by types of adverse events. Subsequently, Medicare extra payment for other types of events might be very different from the reported estimates. In addition, the payment calculations were not exact because some payment adjustments could not be accurately factored in. Third, our findings should be interpreted with care from hospitals’ perspective. Hospitals might be fully paid for the extra costs of treating adverse events occurring in patients covered by fee-for-service plans or prepaid plans, and it is likely that hospitals shift the uncompensated costs from Medicare patients to other patients. More detailed data and sophisticated design are needed to study how hospitals are affected financially by adverse events, given that they serve all types of patients and are paid under multiple systems. Other quality improvement programs. Recent years have witnessed rapid increases in attempts to implement value- or quality-based purchasing and P4P systems that financially reward high-quality providers and penalize low-quality providers.17 For example, the Deficit Reduction Act of 2005 directed Medicare to reduce payment to hospitals when patients acquires infections during a hospital stay, starting 1 October 2008; and the Minnesota-based HealthPartners announced in November 2004 that it would not pay providers for twenty-seven "never" events designated by the National Quality Forum as events that should absolutely never happen.18 To succeed, such systems must take into account the financial incentives that are already embedded in the current system, strengthening the positive incentives and offsetting the negative ones. OUR STUDY SHOWS that the financial aspects of an adverse event are not always apparent, and robust research is needed to recognize the embedded incentives and support the design of new ones. Although this study does not provide a complete business-case analysis, which should have entailed estimation of investments needed, it illustrates the potential returns to investments for Medicare and Medicare-serving hospitals and provides valuable insights into the complexity of the financial aspects of patient-safety improvement.
Chunliu Zhan (czhan{at}ahrq.gov) is a senior health services researcher at the Agency for Healthcare Research and Quality (AHRQ) in Rockville, Maryland. Bernard Friedman is a senior economist at AHRQ. Andrew Mosso is a senior analyst at Social and Scientific Systems Inc. in Silver Spring, Maryland. Peter Pronovost is a professor of critical care medicine and health policy and is director of the Center for Innovations in Quality Patient Care, School of Medicine, at the Johns Hopkins University in Baltimore, Maryland. The authors thank Marge Keyes and James Battles, Agency for Healthcare Research and Quality (AHRQ), and David Hunt, Centers for Medicare and Medicaid Services (CMS), for useful comments. The authors also acknowledge the thirty-six statewide data organizations that participated in the Nationwide Inpatient Sample (NIS) 2002, from the following states: California, Colorado, Connecticut, Florida, Georgia, Hawaii, Illinois, Iowa, Kansas, Kentucky, Maine, Maryland, Massachusetts, Michigan, Minnesota, Missouri, Nebraska, Nevada, New Jersey, New York, North Carolina, Ohio, Oregon, Pennsylvania, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Utah, Vermont, Virginia, Washington, West Virginia, and Wisconsin. The authors are solely responsible for the contents of this paper; no statement herein should be construed as an official position of AHRQ, the U.S. Department of Health and Human Services, or the Johns Hopkins University.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Study Data And Methods
Study Results
