Hospitalist Workload Linked to Costs, Length of Stay

Veronica Hackethal, MD

March 31, 2014

Increased hospitalist workload is linked to higher costs and longer length of stay (LOS), according to a study by researchers at the Christiana Care System in Delaware and Northwestern University in Illinois, published online March 31 in JAMA Internal Medicine.

"This retrospective cohort study of more than 20,000 hospitalizations provides evidence that increased hospitalist workload is associated with reduced efficiency and higher costs," Daniel J. Elliott, MD, MSCE, from the Department of Medicine, Christiana Care Health System, and the Christiana Care Value Institute, Newark, Delaware, and colleagues write. "[T]he magnitude of the difference in LOS and admissions — as much as 2 days and between $5000 and $7500 — is clinically meaningful to hospitals, administrators, and payers."

A census value of 15 might be the tipping point, the researchers add.

As one of the fastest growing medical specialties in the United States, hospitalist medicine has been linked to more efficient, less costly, and better-quality care, according to background information in the article. Hospitalists are under pressure, however, to increase productivity in the face of declining revenues and increased demands for care.

The researchers analyzed inpatient admissions to a private hospitalist service at Christiana Health Care System, which handles more than 75% of hospitalizations in northern Delaware, from February 1, 2008, to January 31, 2011. They adjusted results for patient-, visit-, and hospital-related variables. The researchers categorized hospital occupancy as low (<75%), medium (75% - 85%), or high (>85%).

The analysis included 20,241 hospital admissions involving 13,916 patients. Hospitalists had a mean of 15.5 patient encounters per day.

LOS increased along with workload. For hospitals with less than 75% occupancy, LOS increased linearly from 5.5 days to 7.5 days across low to high workload levels. For occupancies between 75% and 85%, LOS was stable with lower workloads but increased exponentially to about 8.0 days with higher workloads, with the transition occurring at a census value of about 15. For occupancies higher than 85%, LOS decreased slightly in the midrange of workload levels and then significantly increased after that, with the transition occurring again at about a value of 15.

Cost significantly increased as workload and occupancy increased (P < .001 for both). Costs increased by $262 for each additional patient and by $1634 going up each occupancy category. The authors note that LOS was the strongest predictor of cost, but workload remained a significant factor, even after adjustment for LOS.

Workload was not associated with rapid response activation, hospital mortality, patient satisfaction, or readmission.

Limitations included failure to account for physician variability, but the authors point out that the calculated variance between physicians was small and likely did not affect results. In addition, the rate of adverse events may have been underestimated, as the study included only patients at the beginning and end of admission to hospitalist services and may have excluded transfers to other services.

Heavy workloads could affect hospitalists' ability to assess patients, according to the authors, and force them to spend less time on tasks such as documentation and care coordination, which could affect quality of care, delay discharge, and increase costs.

"Although our findings require validation in different clinical settings given the likely variability of these associations across systems, our results suggest that incentives aimed at increasing workload may lead to inefficient and costly care," the authors conclude. "In systems that incentivize physicians based on productivity, consideration should be given to including measures of efficiency and quality."

In an invited commentary, Robert. M. Wachter, MD, from the Department of Medicine, University of California, San Francisco, notes that this study is among the first to look at optimal workload levels for physicians. The results are reassuring, according to Dr. Wachter, because the study did not find higher census levels to be linked to patient harm in terms of rapid response team activation, readmission rates, or hospital mortality.

"[T]his study illustrates that, although 15 patients per hospitalist might not be a magic number in every setting, programs that generally run censuses of more than 15 may want to find ways to lower this workload," Dr. Wachter points out, "The right census number will be the one in a given setting that maximizes patient (and, in a teaching hospital, educational) outcomes, efficiency, and the satisfaction of both patients and clinicians, and does so in an economically sustainable way."

This work was supported by internal funding from the Chairs Leadership Council at Christiana Care Health System. One coauthor has reported serving as division chief at Christiana Care Health System, president and managing physician of the Christiana Medical Group, and consultant to IPC, The Hospitalist Co, Inc. The other authors have disclosed no relevant financial relationships. Dr. Wachter reports serving as past chair and current member of the American Board of Internal Medicine; receiving a contract from the Agency for Healthcare Research and Quality, receiving compensation from John Wiley and Sons, royalties from Lippincott Williams & Wilkins, and McGraw-Hill, and a stipend and stock/options from IPC; serving on the scientific advisory boards for SmartDose, Patient Safety Solutions, CRISI Medical Systems, and EarlySense; and being on the board of directors for Salem Hospital in Oregon.

JAMA Intern Med. Published online March 31, 2014.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.