Embargoed Until:
February 12, 2008
12:00 a.m. Eastern Time



Christopher Fleming

Variations In Hospital Treatment Intensity Are Important Across Different Patient Groups, Researchers Find In California Study

Researchers Find That Hospitals Where Medicare FFS Patients Have Many End-Of-Life Admissions Also Often Admit Other Patients At High Rates

Bethesda, MD -- Large hospital-by-hospital variations in the intensity with which hospitals treat their seriously ill patients is found not only for patients covered by fee-for-service (FFS) Medicare, but also for those covered by Medicare HMOs and for nonelderly patients with private insurance. Moreover, hospitals that frequently hospitalize their FFS Medicare patients are also often the same hospitals that have high levels of resource use for other types of patients as well, according to a study of treatment provided by California hospitals to patients with chronic illnesses in their last two years life. The study, supported by the California HealthCare Foundation, was published today on the Health Affairs Web site. http://content.healthaffairs.org/cgi/content/abstract/hlthaff.27.2.w123

“Earlier research has established that some hospitals provide significantly more resource-intensive treatment to Medicare FFS beneficiaries than other hospitals, and that, on average, higher-intensity hospitals do not achieve better outcomes. In this study, we find similar variations in different age groups and for private insurers in addition to Medicare. Across different age groups and insurers, patients at some hospitals spent many more days in the hospital in their last two years of life than similar patients at other hospitals. This means that there could be important savings available from improving health care delivery for a range of populations,” said lead author Laurence Baker, a professor of health research and policy at Stanford University.

“There is also a tendency for hospitals that provide a resource-intensive approach to treatment for patients in one group to also provide more resource-intensive treatments to other patient groups. Patterns that have been identified in previous analyses focused on Medicare FFS populations, which suggest the potential for substantial savings from moving towards better and more efficient care, thus appear to provide a valuable guide to opportunities available for improvements in other age groups and insurers as well,” said Elliott Fisher of Dartmouth, one of the coauthors of the study.

To measure intensity of treatment, Baker, Fisher, and their coauthor John Wennberg of Dartmouth looked at the total number of inpatient days for which patients were hospitalized in their last two years of life. That measure was further broken down for each patient into the number of hospital admissions and discharges and the average number hospital days per discharge. The researchers examined patients covered by Medicare FFS, Medicare HMOs, private preferred provider organizations (PPOs) and FFS insurers, and private HMOs.

Using data collected by the California state government, Baker and his colleagues looked at all patients who died in California between 1 January 1999, and 31 December 2003 and who had at least one stay in a general acute care hospital during their last two years of life. Patients who visited more than one hospital during this period were assigned to the hospital where they had the greatest number of stays with a medical (as opposed to surgical) diagnosis-related group, or DRG (DRGs are Medicare’s inpatient diagnosis categories). Among the patients studied, the average “loyalty rate” -- the percentage of total hospital days they spent in their assigned hospital -- was greater than 90 percent.

Hospital Characteristics Appear To Drive The Frequency Of Inpatient Admission,
While Coverage Characteristics Are Crucial To The Length Of Each Hospital Stay

In line with earlier research, Baker, Fisher, and Wennberg observed wide variation between hospitals in the intensity of treatment provided to Medicare FFS beneficiaries. There was a fivefold difference between the average number of days used by patients associated with the highest-use hospital (47) and the average number of days used by patients associated with the lowest-use hospital (9). Sizable variations in treatment intensity also existed among Medicare HMO, private PPO/FFS, and private HMO patients. In all three of these insurance groupings, there was at least a twofold difference in total days used by patients associated with the highest- and lowest-use hospitals.

Moreover, hospitals with Medicare FFS patients that used high numbers of hospital days in the last two years of life also tended to have patients using high number of days in the Medicare HMO, private PPO/FFS, and private HMO populations. However, these relationships tracked more closely within insurance types than across them. Thus, Medicare FFS treatment intensity tracked closely with private PPO/FFS treatment intensity, but only moderately with treatment intensity for Medicare HMO and private HMO patients.

The relative influences of hospital and insurance type varied between the two components of treatment intensity: the number of hospital discharges and the average number of hospital days per discharge. The number of discharges was well correlated across all four patient groups: hospitals with high admission rates for Medicare FFS patients also had high admission rates for Medicare HMO, private PPO/FFS, and private HMO patients. “The fact that hospitals with high Medicare FFS discharge rates also had high discharge rates for the other three patient populations suggests that discharge rates in the different insurance groups are all influenced by one or more hospital-specific factors. The most likely candidate is the number of beds at the hospital relative to the population served -- prior research has established that differences in hospital use are largely driven by the relative supply of hospital beds rather than demand,” said Fisher.

On the other hand, the average number of inpatient days per discharge was correlated strongly within insurance types but only weakly across them. Thus, hospitals with high numbers of days per discharge for Medicare FFS patients tended to have high rates of days per discharge for private PPO/FFS patients, but not necessarily for Medicare HMO or private HMO patients. “This suggests that insurance-specific factors rather may be more important than hospital-specific factors in driving the average number of days per discharge. For example, different health plans might vary in their payment methodologies or efforts to promote early discharge,” said Baker.

The work by Baker, Fisher, and Wennberg clarifies the roles that hospitals and health plans play in driving treatment patterns. “The marked differences in admission rates across hospitals and the finding that those with high rates for Medicare FFS patients also have high rates for other payers suggest that both private and public payers could benefit from efforts to align incentives for more efficient use of the hospital as a site of care,” the authors conclude.

After the embargo lifts, you can read the article by Baker, Fisher, and Wennberg athttp://content.healthaffairs.org/cgi/content/abstract/hlthaff.27.2.w123


Health Affairs, published by Project HOPE, is the leading journal of health policy. The peer-reviewed journal appears bimonthly in print with additional online-only papers published weekly as Health Affairs Web Exclusives at www.healthaffairs.org.


©2008 Project HOPE–The People-to-People Health Foundation, Inc.