Trends in market demand for internal medicine 1999 to 2004: an analysis of physician job advertisements

Abstract

Background: The health care marketplace has changed substantially since the last assessment of demand for internal medicine physicians in 1996.

Methods: We reviewed internal medicine employment advertisements published in 4 major medical journals between 1996 and 2004. The number of positions, specialty, and other practice characteristics (e.g., location) were collected from each advertisement.

Results: Four thousand two hundred twenty-four advertisements posted 4,992 positions. Of these positions, jobs in the Northeast (31% of positions) or single specialty groups (36.8% of positions) were most common. The relative proportion of advertisements for nephrologists declined (P < .001), while the relative proportions of advertisements for critical care specialists (0.5% in 1996 to 1.7% in 2004, P = .004) and hospitalists (1.0% in 1996 to 12.1% in 2004, P < .001) increased. Advertisements for outpatient-based generalist positions (i.e., Primary Care and Internal Medicine) declined (-2.7% relative annual change, 95% confidence interval [95% CI] -4.1%, -1.2%) between 1996 and 2004, a decrease largely due to a substantial decline in advertisements noted between 1996 and 1998. However, over the entire time period, the combined proportion of advertisements for all generalists (hospitalists and outpatient-based generalists) did not change (0.5% relative annual change, 95% CI -0.8% to 2.0%).

Conclusions: Since 1996, demand for the majority of medical subspecialties has remained constant while relative demand has decreased for primary care and increased for hospitalists and critical care. Increase in demand for generalist-trained hospitalists appears to have offset falling demand for outpatient generalists.

FIGURE 1

Trends in internal medicine specialty advertisements published between 1996 and 2004, aggregated by procedural (e.g., cardiology, gastroenterology, pulmonary), or nonprocedural (geriatrics, nephrology, rheumatology, hematology, infectious disease) nature of the practice. Critical care is reported separately. Relative to 1996 to 1998, ads for procedural specialties increased in 1999 to 2001, but then declined in 2002 to 2004. Nonprocedural specialties declined early (1996 to 1998 vs 1999 to 2001), but were stable afterward. For discrete time periods, relative changes in critical care advertisements were statistically significant only when 1999 to 2001 was compared with 1996 to 1998. Tests for trends in relative annual changes in advertisements over the entire time period in procedural and nonprocedural specialties are both statistically nonsignificant, while those for critical care are statistically significant (P = .004).

FIGURE 2

Trends in general internal medicine advertisements published between 1996 and 2004, aggregated by site of practice. Early relative increase (1999 to 2001 vs 1996 to 1998) in hospitalist advertisements did not reach tests of statistical significance, but later changes were markedly significant; the reverse trend was observed in outpatient-based generalist advertisements. Hospitalists showed a statistically significant increase over time (P<.0001), while outpatient generalist positions (i.e., Primary care and Internal Medicine) declined over this same time period (P = .001). However, the combined proportion of generalist advertisements (hospitalists and outpatient-based generalists) did not change between 1996 and 2004 (P = .4573).