The Critical Care Workforce:  A Study of the Supply and Demand for Critical Care Physicians

Chapter 4: Comparing Estimates of Supply and Demand

Considerable differences exist between these projections and COMPACCS’ projections of the supply of and demand for intensivists—in part because the current projections model only a subset of the physicians included in the COMPACCS study consistent with a stricter definition of intensivists.  The COMPACCS study used survey data from physicians trained in pulmonology, critical care, or both specialties to determine the number of physicians practicing in an ICU and the average hours per week providing critical care services.  Our analysis relies on the AMA Masterfile to estimate the current intensivist supply, and the AMA data does not allow us to determine which pulmonologists provide critical care in an ICU.  Consequently, the PSM and PDM projections reported here focus on self-designated critical care physicians (with the assumption that all critical care physicians who are active in patient care are providing some services in an ICU).

COMPACCS projections included pulmonologists that care for ICU patients; these physicians tend to be older and are likely to retire from the critical care workforce sooner than their purely intensivist counterparts, thereby projecting a more severe shortage of intensivists.  However, the “effective” supply may be dampened by a decrease in hours worked as has been observed in the medical profession.

Comparing the PSM/PDM and the COMPACCS Projections

The COMPACCS study starts with the assumption that in the base year (1997) intensivist supply and demand are in equilibrium.  This assumption is commonly used in demand/utilization-based forecasting models, but the implication is that the projections are extrapolating year 1997 patterns of care to the future population.  Growth in demand is determined primarily by a growing and aging population.  Thus, the COMPACCS demand projections show relatively little growth until approximately 2010 at which time the size of the elderly population in the U.S. will start to increase dramatically.  The COMPACCS supply estimates are relatively stable during the 30-year projection period.

Although the COMPACCS report was published in JAMA in 2000, data used in the study were from 1997. ^[47]  Since 1997, the percentage of residents choosing to specialize began to change dramatically.  These changes, along with the recent trends in hospital care using more intensivists, illustrate the need for frequent and regular examination of workforce projections.

Why Critical Care Demand Estimates are Unique

The PDM relies on the implicit assumption that physician supply is in balance with physician demand in the base year.  Inefficiencies in the market resulting from any current oversupply or undersupply of physicians will be extrapolated into the future.  Consequently, projections of the future adequacy of supply are relative to recent (i.e., year 2000) conditions and may not account for current unmet demand for services.  In addition, estimates for new or evolving specialties may not fully capture trends in utilization rates, thereby underestimating demand for services.

Critical care is a relatively new specialty and recent growth in intensivist utilization has dramatically outpaced the growth in demand related to a growing and aging population.  If historical utilization rates are extrapolated into the future, then aggregate demand for intensivists does not appear to exceed available supply.  However, recent trends suggest that a growing proportion of critically ill patients will receive intensivist services, so that current utilization and service delivery patterns underestimate the likely current and future demand.

This weakness is especially true in critical care because of the changing nature of delivery and organization of services in the ICU.  It becomes particularly important in analysis of the critical care workforce because of the evidence regarding the current inadequacy of ICU staffing.  The assumption that supply and demand are in equilibrium at baseline cannot be made for critical care practice because intensivists currently care for only one-third of critically ill patients.  Given the level of evidence supporting intensivist-directed care for ICU patients, two-thirds of patients may be receiving less than optimal care.  Even if only half of patients admitted to intensive care units were cared for by full-time intensivists, there would be a shortage of critical care physicians in the range of 25 percent of current supply (Exhibit 15).  This shortage is despite expected modest increases in efficiency of care (i.e., decreased length of stay) for patients cared for by intensivists. ^[48]

Exhibit 15. Projected Supply vs. Optimal Utilization for Intensivists, 2000-2020

[D]

It should be noted that these projections, which assume a current shortage of intensivists, also differ from COMPACCS projections.  The absolute magnitude of shortages remain below the level predicted by COMPACCS because Angus and colleagues utilized survey data to provide estimates of intensivists which incorporated time spent by physicians trained in critical care, pulmonology, or both.  This analysis was based upon a stricter definition of intensivist and included only physicians trained in critical care.  As a result, the COMPACCS study included a greater number of intensivists at baseline.  The larger shortage projected in that study is, in part, due to the fact that pulmonologists tend to be older than their purely intensivist counterparts; a larger proportion of physicians practicing at baseline in the COMPACCS study were expected to retire earlier than expected in our projections.

However, we believe both approaches to be methodologically sound.  Because both sets of projections trend the current supply forward, they express supply (and demand) changes based upon a definition that remains consistent over time.  So, while absolute shortages of intensivists as defined by COMPACCS are difficult to compare with those projected in this study, shortages of intensivists as a proportion of current supply should be comparable to one another.