The financial incentive for hospitals to prevent nosocomial infections under the prospective payment system. An empirical determination from a nationally representative sample

To clarify the financial incentives for hospitals to prevent nosocomial infections, we analyzed 9423 nosocomial infections identified in 169 526 admissions selected randomly from the adult admissions to a random sample of US hospitals. By classifying each admission into a baseline diagnosis related group (DRG) (after first excluding all diagnoses of nosocomial infection) and a final DRG (after including these diagnoses), we found that only 5% to 18% of nosocomial infections would have caused the admission to be reclassified to a higher-paying DRG, depending on the extent to which physicians recorded nosocomial infection diagnoses in patients' medical records. The extra payment from the reclassification, averaged over all nosocomial infections, would have been no more than $93 per infection (in 1985 reimbursement rates), constituting only 5% of the hospitals' costs for treating these infections. Thus, at least 95% of the cost savings obtained from preventing nosocomial infections represents financial gains to the hospital.     

The relationship between hospital per diem billing and DRG reimbursement for urban trauma patients

Study Objective: To study the relationship between a trauma center per diem charges and medicare DRG reimbursement. Design: Retrospective comparison of charges ($630/day, $1500/ICU day) and hypothetical DRG reimbursement using medical records ICD-9 N and P codes and version 5.0 of grouper. Setting: An urban level I trauma center that participates in a trauma system that serves a population of 3 million people. Patient Population: Trauma patients ≥16 years old (mean age of 32 years) admitted and discharged between 1/1/88 and 9/30/88. The group was 86% male, 75% black, with a blunt mechanism of injury in 64%. The mean ICU stay was 0.9 days, and the mean total length of stay was 5.0 days. Results: Total per diem charges were $8,652,159, and DRG reimbursement was $8,636,505, causing a net loss of $15,654, or 0.2% of charges. Mean charges and reimbursement did not differ for the entire group. The mean loss per patient was $8. Mean charges and reimbursement differed in penetrating trauma patients (mean loss=$138), as well as those with different lengths of stay. The correlation between charges and reimbursement was 0.42; for penetrating trauma patients, the correlation was 0.58. (p<.001) Conclusion: If DRG reimbursement were provided for all admitted trauma patients, the amount would equal per diem rates. Trauma centers with similar patients and lengths of stay can use these per diem rates to estimate DRG reimbursement.     

The relationship between hospital per diem billing and DRG reimbursement for urban trauma patients

Study Objective: To study the relationship between a trauma center per diem charges and medicare DRG reimbursement. Design: Retrospective comparison of charges ($630/day, $1500/ICU day) and hypothetical DRG reimbursement using medical records ICD-9 N and P codes and version 5.0 of grouper. Setting: An urban level I trauma center that participates in a trauma system that serves a population of three million people. Patient Population: Trauma patients ≥16 years old (mean age of 32 years) admitted and discharged between 1/1/88 and 9/30/88. The group was 86% male, 75% black, with a blunt mechanism of injury in 64%. The mean ICU stay was 0.9 days, and the mean total length of stay was 5.0 days. Results: Total per diem charges were $8,652,159, and DRG reimbursement was $8,636,505, causing a net loss of $15,654, or 0.2% of charges. Mean charges and reimbursement did not differ for the entire group. The mean loss per patient was $8. Mean charges and reimbursement differed in penetrating trauma patients (mean loss=$138), as well as those with different lengths of stay. The correlation between charges and reimbursement was 0.42; for penetrating trauma patients, the correlation was 0.58. (p<.001) Conclusion: If DRG reimbursement were provided for all admitted trauma patients, the amount would equal per diem rates. Trauma centers with similar patients and lengths of stay can use these per diem rates to estimate DRG reimbursement.     

Medicare reimbursement accuracy under the prospective payment system, 1985 to 1988.

BACKGROUND--Hospital reimbursement by Medicare's prospective payment system depends on accurate identification and coding of inpatients' diagnoses and procedures using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). A previous study showed that 20.8% +/- 0.5% (mean +/- SE) of hospital bills for 1985 contained errors that changed their diagnosis related group (DRG) and that a significant 61.6% +/- 1.3% of errors overreimbursed the hospitals. This DRG "creep" improperly increased net reimbursement by 1.9%, +308 million when projected nationally. The present study updated our previous study with 1988 data. METHODS--The Office of Inspector General, US Department of Health and Human Services, obtained a simple random sample of 2451 hospital charts for Medicare discharges from 1988. The American Medical Record Association reabstracted the ICD-9-CM codes on a blinded basis, grouped them to DRGs, and determined the reasons for discrepancies. RESULTS--Coding errors declined to 14.7% +/- 0.7% in 1988, and a nonsignificant 50.7% +/- 2.6% of DRG errors overreimbursed the hospitals. Projected nationally, hospitals did not receive a significant overreimbursement. Physician misspecification of the narrative diagnoses underreimbursed the hospitals, while billing department resequencing overreimbursed them. CONCLUSIONS--The attestation requirement may have deterred DRG creep due to attending physician upcoding, but the peer review organizations' sentinel effect and educational activities have not eliminated hospital resequencing.     

The emergency department as a pathway to admission for poor and high-cost patients

BACKGROUND.--To determine the importance of the emergency department as the means of access to the hospital for the poor and the fiscal implications of providing these services, we examined the relationship between patients' socioeconomic status and admission via the emergency department. We also determined the association between entering the hospital via the emergency department and hospital resource use. METHODS.--We conducted a study of 20,089 patients admitted to five Massachusetts hospitals (three community, two tertiary care) during a 6-month period. We determined the proportions of patients within various socioeconomic and disease groupings who entered through the emergency department. We compared length of stay and charges for patients admitted through the emergency department with those for patients admitted through other routes. RESULTS.--Overall, 51% of patients entered via the emergency department. Elderly patients (age greater than 65 years; odds ratio, 1.87) and patients with lower socioeconomic status as measured by income, occupation, and education (odds ratios, 2.38, 1.47, and 1.69, respectively) were more likely to enter the hospital via the emergency department than other patients. After adjustment for diagnosis related group, severity as measured by DRGSCALE, and socioeconomic status as measured by income, and excluding outliers, patients admitted via the emergency department stayed 27% longer and incurred 13% higher charges than other patients (P less than .001). CONCLUSIONS.--Our data indicate that patients with lower socioeconomic status are more likely than other patients to use the emergency department as their means of access to the hospital and that patients admitted via the emergency department use far more resources than patients in the same diagnosis related group admitted by other means. Hospitals that make emergency department services more available may be more likely to hospitalize socioeconomically disadvantaged patients and may be at a substantial financial disadvantage under per-case reimbursement systems such as Medicare.     

A New Accounting System for Financial Balance Based on Personnel Cost After the Introduction of a DPC/DRG System

A hospital director must estimate the revenues and expenses not only in a hospital but also in each clinical division to determine the proper management strategy. A new prospective payment system based on the Diagnosis Procedure Combination (DPC/PPS) introduced in 2003 has made the attribution of revenues and expenses for each clinical department very complicated because of the intricate involvement between the overall or blanket component and a fee-for service (FFS). Few reports have so far presented a programmatic method for the calculation of medical costs and financial balance. A simple method has been devised, based on personnel cost, for calculating medical costs and financial balance. Using this method, one individual was able to complete the calculations for a hospital which contains 535 beds and 16 clinics, without using the central hospital computer system.     

The 1987 US hospital AIDS survey

In 1987, the National Public Health and Hospital Institute conducted a national survey of 623 acute-care hospitals to obtain information relating to inpatient and outpatient care for persons with acquired immunodeficiency syndrome (AIDS). Two hundred seventy-six hospitals reported treating persons with AIDS; the average length of stay was 16.8 days. Average costs and revenues per inpatient day were $681 and $545, respectively, with a cost per patient per year of $17,910. Estimated cost for AIDS inpatient care during 1987 was $486 million; Medicaid represented the primary payer. Regional and ownership comparisons for this year and between 1985 and 1987 indicated significant differences in utilization, payer source, and financing. Results suggest major differences in reimbursement and losses related to payer source or lack of insurance, with many hospitals that serve large numbers of low-income persons with AIDS encountering moderate to severe financial shortfalls. We conclude that increasing concentrations of persons with AIDS in relatively few hospitals in large cities may make it more difficult to secure the broader political base necessary to obtain adequate support for treatment.     

Expanding Medicare and employer plans to achieve universal health insurance

This article presents a proposal for expanding Medicare and employer-based health insurance plans to achieve universal health insurance. Under this proposed health care financing system, employees would provide basic health insurance coverage to workers and dependents, or pay a payroll tax contribution toward the cost of their coverage under Medicare. States would have the option of buying all Medicaid beneficiaries and other poor individuals into Medicare by paying the Medicare premiums and cost sharing. Other uninsured individuals would be automatically covered by Medicare. Employer plans would incorporate Medicare's provider payment methods. This proposal would result in incremental federal governmental outlays on the order of $25 billion annually. These new federal budgetary costs would be met through a combination of premiums, employer payroll tax, income tax, and general tax revenues. The principal advantage of this plan is that it draws on the strengths of the current system while simplifying the benefit and provider payment structure and instituting innovations to promote efficiency.     

A critical analysis of the resource-based relative value scale.

BACKGROUND--There is a general perception that procedural medical services are reimbursed at an inappropriately greater rate than cognitive services. By congressional mandate, the Health Care Financing Administration (HCFA) has been directed to establish a Medicare fee schedule to shift funding under a budget-neutral assumption from procedural to cognitive services. To provide a rational basis for this change, Hsiao et al (Harvard-Hsiao) developed a resource-based relative value scale (RBRVS) that equates the value of a service to the resources necessary to generate the service. METHODS--Instead of focusing on relative values and fee schedules ("price-per-unit-service"), the present study employs the standard commercial/industrial method of determining reimbursement rate (income divided by hours of labor) for 15 medical and surgical specialties. Data from independent sources are used to determine income and hours of professional effort for each of the specialties studied. Harvard-Hsiao and HCFA predicted the percent change in income for each of the specialties under the initial RBRVS and the HCFA fee schedule. The predicted income was then employed in this study to recompute reimbursement rates under the newly proposed payment systems. RESULTS: CURRENT PAYMENT SYSTEM--Average annual incomes for medical and surgical specialties are $124,500 and $176,600, respectively, a 42% difference (P = .03). Average weekly work hours (nominal hours, as adjusted for overtime) for medical and surgical specialties are 70.6 and 87.8, respectively (P = .005). Average hourly reimbursement rates for medical ($33.90) and surgical ($38.80) specialties are not substantially different (P, not significant). The difference in annual income is explained by the 17.2 hours per week of additional work hours by surgeons. The erroneous perception that procedurists are reimbursed at a higher rate than cognitive practitioners likely arises from differences in billing methods by which surgeons shift charges for cognitive work hours to the 18% of their time spent in the operating room. RESULTS: PROPOSED RBRVS AND HCFA PAYMENT SYSTEM--The income of all specialties is equalized about a mean of $132,500 (+/- $21,400 [1 SD]) by varying reimbursement rates in such a way that the effect of working hours is fully discounted. Reimbursement rates under the proposed payment system make no recognition of the hours of professional effort, postgraduate specialty training, or putative differences in the nature of the physician's work. CONCLUSION.--The RBRVS, and the HCFA fee schedule to the extent that it is based on that scale, are inappropriate bases for the reform of the physician reimbursement system.     

Costs, charges and revenues of elective coronary angioplasty and stenting: the public versus the private system

OBJECTIVE: To calculate the costs of elective coronary angioplasty and stenting (CAS) in the public and private healthcare systems and to compare these costs with the charges levied and the revenues obtained. DESIGN: A prospective health economics study. SETTING: A tertiary care public hospital and a co-located tertiary care private hospital in the 12 months from February 1998. STUDY POPULATION: 186 consecutive patients (124 public, 62 private) undergoing elective CAS. MAIN OUTCOME MEASURES: Outcome of CAS; exact costs of CAS in the two hospitals; exact charges to private patients; estimated charges in a typical, not co-located, "industry standard private hospital"; estimated costs to the Federal Government of CAS in the public and private system. RESULTS: The immediate and six-month outcomes in the two groups were similar. The average cost of CAS in public patients was $5,516, compared with $5,844 in private patients. The length of stay, number of stents per case and use of nonstent consumables was similar for both groups. Average charges for CAS in patients in the co-located private hospital were $13,347, and estimated average charges for CAS in an industry standard private hospital were $14,978. Estimated current costs to the government for CAS in a public hospital, a co-located private hospital, and an industry standard private hospital were $5664, $5,394 and $6,201, respectively. CONCLUSIONS: Despite similar treatments and similar treatment costs, CAS in the private system, as a consequence of the charges levied, is more than twice as expensive as in the public system, with government costs similar for both systems. These data (together with data from other studies showing that CAS is performed more frequently in private patients) suggest that encouraging more people to take out private health insurance will, paradoxically, increase government costs for CAS as well as increasing overall health expenditure.     

Impact of the Medicare fee schedule on an academic department of medicine.

OBJECTIVE--To examine the effect of the Medicare Fee Schedule (MFS) on Medicare revenues in the department of medicine at an urban academic medical center after the MFS is fully implemented. METHODS--Department revenues from Medicare were compared with projected revenues using the MFS proposed by the Health Care Financing Administration on June 5, 1991. National Medicare claims data were used to determine differences in service mix between community and academic internists and the impact of the geographic component of the MFS on department revenues. RESULTS--Department revenues from Medicare in 1996 are projected to be 25.5% lower under the MFS than if the current system had continued. Subspecialty sections that perform large numbers of procedures and special tests had the largest decrease in revenues (eg, gastroenterology, -29.8%); however, this did not differ greatly from decreases in sections that mainly provide visits and consultations (eg, general internal medicine, -24.7%). CONCLUSION--The proposed MFS is projected to lead to substantial reductions in department revenues from Medicare. While relative values for services and geographic location will play a role in how individual departments fare under the MFS, the value of the conversion factor used in the final MFS will be the factor of greatest importance.     

Medicare risk contracting. Lessons from an unsuccessful demonstration

The Tax Equity and Fiscal Responsibility Act of 1982 provided a full-risk Medicare capitation financing option for health maintenance organizations and competitive medical plans. Two rounds of demonstrations were conducted, followed by the publication of final regulations in January 1985. The first-round demonstration at Marshfield, Wis, was operational for 28 months. Thirty-seven percent of all resident beneficiaries enrolled. Aggregate losses exceeded $3 million (11.6% of revenue). Management implemented increasingly more stringent utilization review. Overall hospital utilization declined 261.7 days per 1000 from fiscal year 1981 to 1982; nonetheless, federal reimbursement was insufficient to meet program costs and the demonstration was terminated. The central reimbursement method used in Medicare risk contracting (adjusted average per capita cost) does not adequately control for enrollment selection, unmet medical need, or recent regional cost variations. Reimbursement set at 95% of estimated fee-for-service costs does not recognize, and in the long run will not support, an efficiently operating delivery system.     

The economic consequences of medical injuries. Implications for a no-fault insurance plan.

BACKGROUND--There has been little research into the actual economic consequences of medical injuries. This inhibits informed discussion of alternatives to malpractice litigation. For example, the cost of no-fault medical accident insurance has been thought to be prohibitive. METHOD--As part of a comprehensive analysis of medical injury and litigation, we interviewed a random sample of 794 individuals who had suffered medical adverse events in New York hospitals in 1984 and used their responses to calculate the cost of injuries. We then estimated the costs of a simulated no-fault insurance program that would operate as a second payer to direct insurance sources and would compensate for all financial losses attributed to medical injury. RESULTS--The estimated costs that would be paid by a simulated no-fault program were $161 million for medical care, $276 million for lost wages, and $441 million in lost household production, or a total of $878 million in 1989 dollars for the cohort of patients who were injured in 1984. CONCLUSION--Although our estimate does not include administrative costs, it nonetheless indicates that a no-fault program would not be notably costlier than the more than $1 billion New York physicians now spend annually on malpractice insurance.     

Cost of and payment source for pelvic inflammatory disease. Trends and projections, 1983 through 2000

Pelvic inflammatory disease (PID) and its sequelae affect millions of women in the United States at substantial costs. To estimate these total costs annually and to determine payment sources, we analyzed data from local, state, and national sources. Direct costs for PID and PID-associated ectopic pregnancy and infertility were estimated to be $2.7 billion, and indirect costs were estimated to be $1.5 billion, for a total cost of $4.2 billion in 1990. Overall, private insurance covered the largest portion of the direct costs of PID (41%), followed by public payment sources (30%). However, the proportion of payments made by private insurance appears to be decreasing, while that by public payment sources is increasing. In the year 2000, costs associated with PID are projected to approach $10 billion if the current PID incidence persists, with an increasing proportion of this expense burdening public institutions. Prevention of PID is needed both to reduce human suffering and to contain rising costs.     

Severity of illness and the teaching hospital

In the current environment of cost containment pressures on health care providers, teaching hospitals are facing increased financial risks that could jeopardize their special role in the health care delivery system. One of these risks is that the Medicare prospective payment system does not adequately account for severity of illness. Whether teaching hospitals treat a case mix of patients with more severe illness than do nonteaching hospitals was tested in the study reported here using two severity measures, Horn's severity of illness index and Gonnella's "disease staging." Teaching hospitals were found to treat a significantly greater proportion of severely ill patients than community hospitals, especially when measured by the severity of illness index. Differences in case mix of severity of illness among hospitals can have a significant impact on patient care costs, which may not be adequately met by a reimbursement system based on diagnosis related groups. Hospital managers can use severity of illness measures to assess the resource needs of patients and the practice patterns of physicians. If severity of illness measures help describe the special burden of treatment that teaching hospitals bear, they should be used to establish the case for adequate financial support.     

Comparing outcomes of care before and after implementation of the DRG-based prospective payment system

We compared patient outcomes before and after the introduction of the diagnosis related groups (DRG)-based prospective payment system (PPS) in a nationally representative sample of 14,012 Medicare patients hospitalized in 1981 through 1982 and 1985 through 1986 with one of five diseases. For the five diseases combined; length of stay dropped 24% and in-hospital mortality declined from 16.1% to 12.6% after the PPS was introduced (P less than .05). Thirty-day mortality adjusted for sickness at admission was 1.1% lower than before (16.5% pre-PPS, 15.4% post-PPS; P less than .05), and 180-day adjusted mortality was essentially unchanged at 29.6% pre-vs 29.0% post-PPS (P less than .05). For patients admitted to the hospital from home, 4% more patients were not discharged home post-PPS than pre-PPS (P less than .05), and an additional 1% of patients had prolonged nursing home stays (P less than .05). The introduction of the PPS was not associated with a worsening of outcome for hospitalized Medicare patients. However, because our post-PPS data are from 1985 and 1986, we recommend that clinical monitoring be maintained to ensure that changes in prospective payment do not negatively affect patient outcome.     

Hospital behavior under competition and cost-containment policies. The California experience, 1980 to 1985

Previous studies of hospital competition have found that greater competition leads to higher hospital costs. We describe herein the change in behavior of California's hospitals since the introduction of competitive and cost-containment programs. To examine the impact of California's pro-competition policies on hospital performance, we grouped the state's short-term hospitals according to the level of competition within their markets. From 1983 through 1985, total inpatient costs (inflation adjusted) increased by less than 1% in hospitals in low-competition markets compared with a decrease of 11.29% in hospitals located in highly competitive markets. After controlling for the effects of the Medicare prospective payment system program, the rate of increase in cost per discharge in hospitals in highly competitive markets was 3.53% lower than the rate of increase in hospitals in low-competition markets during the period from 1983 through 1985. We conclude that these pro-competition policies are having dramatic and potentially far-reaching effects on the nature of hospital competition, leading to increased competition based on price.     

Medical care costs of patients with AIDS in San Francisco

This article reports on the findings of a study of medical care expenditures of persons with acquired immunodeficiency syndrome (AIDS) treated at San Francisco General Hospital in 1984. We found that mean charges per AIDS hospital admission were $9024, mean charges of patients with AIDS who received all their hospital inpatient and outpatient care at San Francisco General Hospital in 1984 ranged from $7026 to $23,425, and mean lifetime inpatient charges of patients with AIDS who died and who had received all their inpatient care at the hospital were $27,571. These latter charges were considerably lower than previously published estimates of lifetime direct medical care costs of patients with AIDS, and the possible reasons for the differences include much lower lifetime use of hospital services and somewhat lower cost per hospital day.     

Potential cost to Western Australia of proposed patient co-payments according to healthcare organisational structure: A preliminary analysis

Background

The Australian federal government has proposed an AUD $7 patient co-payment for a general practitioner (GP) consultation. One effect of the co-payment may be that patients will seek assistance at public hospital emergency departments (EDs), where currently there is no user charge.

Aims

We studied the possible financial impact of patient diversion on the Western Australia (WA) health budget.

Method

We constructed a spreadsheet model of changes in annual cash flows including the co-payment, GP fees for service, and rates of diversion to emergency departments with additional marginal costs for ED attendance.

Results

Changes in WA cash flows are the aggregate of marginal ED costs of treating diverted patients and added expenditure in fees paid to rural doctors who also man local emergency centres. The estimated costs to WA are AUD $6.3 million, $35.9 million and $87.4 million at 1, 5, and 10 per cent diversion, respectively. Commonwealth receipts increase and expenditure on Medicare benefits declines.

Conclusion

A diversion of patients from GP surgeries to ED in WA caused by the co-payment will result in increased costs to the state, which may be substantial, and will reduce net costs to the Commonwealth.     

Impact of the Medicare fee schedule on payments to physicians

Beginning in 1992, the Medicare program will pay physicians by the Medicare Fee Schedule, a system of geographically adjusted standardized payment rates based in part on the Resource-Based Relative Value Scale developed by Hsaio et al and in part on current Medicare payments. In our simulations of the Medicare Fee Schedule, we find that (1) redistributions of Medicare-allowed charges across specialities will be substantial but approximately only half the size projected by Hsaio, (2) there will be large redistributions among geographic areas that tend to compound the specialty redistributions, and (3) there will be wide variation within specialties as to how individual providers are affected. The majority of the redistributive impact of the Medicare Fee Schedule is attributable to implementation of a geographically adjusted system of standardized payments rather than to the particular work values developed by Hsiao et al in the Resource-Based Relative Value Scale.