Growth in the care of older patients by hospitalists in the United States

Abstract

Background: National and population-based information on the increase in patient care by hospitalists in the United States is lacking.

Methods: Using a 5% sample of Medicare beneficiaries in 1995, 1997, 1999, and the period from 2001 through 2006, we identified 120,226 physicians in general internal medicine who were providing care to older patients in 5800 U.S. hospitals. We defined hospitalists as general internists who derived 90% or more of their Medicare claims for evaluation-and-management services from the care of hospitalized patients. We then calculated the percentage of all inpatient Medicare services provided by hospitalists and identified patient and hospital characteristics associated with the receipt of hospitalist services.

Results: The percentage of physicians in general internal medicine who were identified as hospitalists increased from 5.9% in 1995 to 19.0% in 2006, and the percentage of all claims for inpatient evaluation-and-management services by general internists that were attributed to hospitalists increased from 9.1% to 37.1% during this same period. Accompanying the increase in care by hospitalists was an increase in the percentage of all hospitalized Medicare patients who were treated by general internists (both hospitalists and traditional, non-hospital-based general internists), from 46.4% in 1995 to 61.0% in 2006. In a multilevel, multivariable analysis controlling for patient and hospital characteristics, the odds of receiving care from a hospitalist increased by 29.2% per year from 1997 through 2006. In 2006, there was marked geographic variation in the rates of care provided by hospitalists, with rates of more than 70% in some hospital-referral regions.

Conclusions: These analyses of data from Medicare claims showed a substantial increase in the care of hospitalized patients by hospitalist physicians from 1995 to 2006.

Figure 1. Distribution of Physicians in General Internal Medicine According to the Percentage of Evaluation-and-Management Medicare Claims Generated from the Care of Hospitalized Patients, 2001 through 2006

In 2001, 30.0% of general internists derived less than 10% of their evaluation-and-management billing codes from the care of hospital inpatients, and 9.7% of general internists derived 90% or more of these billing codes from the care of hospital inpatients. From 2001 through 2006, the percentage of physicians generating less than 10% or 90% or more of these billing codes from the care of hospital inpatients increased. Conversely, the percentage of physicians with more balanced distributions between inpatient and outpatient services decreased.

Figure 2. Growth of Hospitalist Care in U.S. Hospitals from 1995 through 2006

Panel A shows the percentage of hospitalists over time, according to medical specialty. The change over time in the percentage of Medicare admissions involving care by physicians in general internal medicine who received all of their general internal-medical care from hospitalists is shown, stratified according to geographic region (Panel B), teaching status of the hospital (Panel C), hospital size (Panel D), the patient’s race or ethnic group (Panel E), type of admission (emergency vs. nonemergency) (Panel F), and type of surgical (Panel G) or medical (Panel H) diagnosis. Rapid growth of hospitalist care occurred in all patient and hospital groups and in all geographic regions. DRG denotes diagnosis-related group.

Figure 3. Hospitalist Care of Medicare Patients in 306 Hospital-Referral Regions in 1997, 2001, and 2006

The maps show the percentage of Medicare patients, among those cared for by physicians in general internal medicine, who received all their care from hospitalists, according to the hospital-referral region. In 1997, there was little variation among regions; in 85.0% of the hospital-referral regions, less than 10% of hospitalized Medicare patients received care from a hospitalist. Both the percentage of patients receiving hospitalist care and the variation among hospital-referral regions increased from 1997 to 2001 and from 2001 to 2006. White areas on the maps indicate areas without residents, such as national parks.

Figure 4. Pattern of Medicare Claims in the Years before and after 2002 among Physicians in General Internal Medicine Who Were Identified as Hospitalists in 2002

In 2002, a total of 6898 physicians in general internal medicine generated 90% or more of their Medicare claims from services to hospitalized patients. The bars show the percentage of these physicians who were predominantly practicing hospital medicine, predominantly practicing outpatient medicine, or generating few or no Medicare claims in the years before and after 2002. Only 36.9% of these physicians devoted the majority of their time to inpatient care in 1999; 9.5% were primarily outpatient physicians, 2.9% were listed in other specialties, and 50.7% had few or no billing cades (indicating that they were still in training). By 2006, a total of 63.0% of the physicians identified as hospitalists in 2002 were still devoting the majority of their time to inpatients, 8.8% were primarily outpatient general internists, 9.2% were in other specialties, and 19.0% had few or no evaluation-and-management charges (suggesting that they might have returned to training). In these analyses, “predominantly practicing hospital or outpatient medicine” was defined as 50% or more of evaluation-and-management claims generated from the care of hospital patients or outpatients, respectively. “Few or no Medicare claim” was defined as fewer than five claims per year. Data for the years 1998 and 2000 were not available