Exploring the Potential Causes of the Emergence of Hospitalists: Chicken vs. Egg
- First Online:
- Cite this article as:
- Leykum, L. & Mortensen, E. J GEN INTERN MED (2010) 25: 378. doi:10.1007/s11606-010-1320-2
With the introduction of the term “hospitalist” by Wachter and Goldman1 in 1996 hospital medicine has expanded exponentially2,3. Since that time the odds that an inpatient would receive care from a hospitalist, rather than their primary care physician, has grown by almost 30% per year2. In 2006 a study of Medicare recipients revealed that almost 20% of all general internists identified themselves as hospitalists2. The literature has described the trajectory of this growth1–3 and its impact on the quality of care received3–5 in considerable detail. However, data related to the forces behind this growth has been much less plentiful6,7.
Economic, efficiency, and quality pressures have all been considered as factors in the growth of the hospitalist model of care1,6,8. Explanations using these factors have included the impacts of managed care on the need for primary care physicians to see more patients in the outpatient setting, and of hospital reimbursement changes leading to both higher acuity inpatients and to pressures to discharge patients as quickly and efficiently as possible. It has been argued that these pressures have combined to make care of hospitalized patients by primary care physicians less feasible and/or desirable, promoting the hospitalist model of care. The hospitalist model allowed primary care physicians to focus on outpatient care, and allowed physicians dedicated to the care of hospitalized patients to provide efficient on-site services for an increasingly sick inpatient population.
In this transition to a hospitalist model, it has not been clear how much growth was driven by a retreat of primary care physicians from hospital care versus their being “pushed” out by increasing numbers of hospitalists. The paper by Drs. Meltzer and Chung9 contributes to our understanding of the growth of the hospitalist movement by looking at this issue.
To examine the question of whether hospitalist growth was influenced primarily by decreasing primary care involvement in the care of hospitalized patients versus increasing hospitalist activity, Drs. Meltzer and Chung use data from several national sources over the time period of 1980 to 2005 to examine the trends in primary care physician hospital activity before and after the inception of the hospitalist model. They use discharge and length of stay data to calculate the total number of inpatient encounters, and divide this by the number of generalist physicians to calculate the “average inpatient encounters relative to the generalist workforce.” They also calculate the probability of a patient being admitted from their primary care provider’s office. Their rationale in looking at these metrics was to understand whether primary care physicians began caring for fewer hospitalized patients prior to the adoption of the hospitalist model, implying that they had fewer numbers of patients and economic or other incentives to continue this practice. Additionally, the trends in primary care presence after the adoption of the hospitalist model would help to answer the question of whether primary care physicians are being “crowded out” by hospitalists.
Their analysis suggests that primary care physicians had declining inpatient activity well before the onset of the hospitalist movement, both in terms of the average number of inpatient encounters and the likelihood that a patient seen in their offices would need to be admitted. Based on these findings, they conclude that disincentives for primary care physicians preceded the hospitalist movement, and that hospitalists have not “crowded out” primary care physicians.
Drs. Meltzer and Chung’s analysis has the key strength of using data from national sources. Many examinations of hospitalist programs to date have been on a small scale, making their generalizability limited6,7. This analysis has the additional strength of including data over a long time period (25 years), allowing them to be more certain that their findings reflect larger trends.
However this study’s aggregate nature may also be considered a major limitation. Due to the use of national data, we are unable to ascertain the specific impacts of different providers, health plans, and /or economic factors at the local or regional level. It may be true that general internists and other primary care physicians were retreating from inpatient care in many rural areas where overall fewer numbers of physicians would decrease the possibility of feeling “crowded out”. In still other areas, physician activity in hospital setting may not have changed significantly, or groups may have assigned individual physicians to rotating hospital coverage. In large health care systems, that included larger numbers of both primary care physicians and specialists, primary care physicians may have experienced more pressure to focus on outpatient care. In still other regions, physician activity in hospital settings may not have changed significantly, or healthcare groups may have assigned individual physicians to rotating hospital coverage to more efficiently care for the smaller number of patients not covered by hospitalists.
In addition, as the authors point out, their findings are suggestive, but do not directly measure the incentives on primary care physicians to care for hospitalized patients. Although primary care physicians seem to have fewer patients in the hospital over time we cannot assume that caring for patients while they were hospitalized was less important or less attractive. It is possible that primary care physicians still felt “crowded out” as they were taking care of fewer inpatients.
The importance of parsing out the root factors leading to the growth of the hospitalist model may be debated. The point has been made that the hospitalist movement “has arrived”10, and that efforts should be targeted to improving coordination of care between primary care and hospitalist physicians and assisting both groups to evolve their practices further. However, in this time of potential changes in the structure of health care and physician reimbursement in both the primary care and hospitalist settings, understanding the widespread impact of financial and other pressures on care delivery will help us to make better choices in improving on what we have. Drs. Meltzer and Chung’s analysis is an important contribution in that regard.