Mia Djulbegovic1,2,3, Kevin Chen1,2,4, Andrew B Cohen2,5, Daniel Heacock6, Maureen Canavan1,7, William Cushing6, Ritu Agarwal5,8, Michael Simonov9, Sarwat I Chaudhry1,5. 1. Department of Internal Medicine, National Clinician Scholars Program, Yale University School of Medicine, New Haven, Connecticut, USA. 2. Veterans Affairs Connecticut Healthcare System, West Haven, Connecticut, USA. 3. Department of Medicine, Division of Hematology and Oncology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA. 4. New York University Grossman School of Medicine, Division of General Internal Medicine and Clinical Innovation, New York, New York, USA. 5. Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut, USA. 6. Yale-New Haven Hospital, New Haven, Connecticut, USA. 7. Department of Internal Medicine, Yale Cancer Outcomes, Public Policy and Effectiveness Research (COPPER) Center, Yale School of Medicine, New Haven, Connecticut, USA. 8. Joint Data Analyst Team, Yale New Haven Health System, New Haven, Connecticut, USA. 9. Department of Internal Medicine, School of Public Health, Yale University, New Haven, Connecticut, USA.
Abstract
BACKGROUND: Hospitalist physicians' workload-the total number of patients they care for daily-is rising in the U.S. Hospitalists report that increased workload negatively affects patients care. OBJECTIVE: Measure the associations between hospitalist physicians' workload and clinical outcomes. DESIGN, SETTINGS, AND PARTICIPANTS: Observational study, using electronic health record (EHR) data, of adults hospitalized on the hospitalist service at Yale-New Haven Hospital from 2015-2018. MAIN OUTCOME AND MEASURES: We defined hospitalists' workload as the number of patients they cared for on the first full hospital day of a given patient's encounter. We used multilevel Poisson and logistic regression to examine associations between workload and length of stay (LOS), return to the Emergency Department (ED), and readmission. We adjusted for sociodemographic factors, patient complexity and severity of illness, and weekend admission (for LOS) or discharge (for ED visits or readmission). RESULTS: We analyzed 38,141 hospitalizations. Median patient age was 64 years (IQR 51-78 years), 53% were female, and 34% were nonwhite. Mean workload was 15 patients (SD 3 patients; range 10-34 patients). LOS was prolonged by 0.05 days (95% CI 0.02, 0.08; p(0.001) when comparing the 75th workload percentile (16 patients) to the 25th workload percentile (13 patients). There were no associations between workload and ED visits or readmission within 7 and 30 days. CONCLUSIONS: There was a statistically significant but modest relationship between workload and LOS; workload was not associated with ED visits or readmissions.Given clinical reports of the deleterious effects of increased hospitalist workload, there is a need for prospective research assessing a range of outcomes, beyond those measurable in contemporary EHR data.
BACKGROUND: Hospitalist physicians' workload-the total number of patients they care for daily-is rising in the U.S. Hospitalists report that increased workload negatively affects patients care. OBJECTIVE: Measure the associations between hospitalist physicians' workload and clinical outcomes. DESIGN, SETTINGS, AND PARTICIPANTS: Observational study, using electronic health record (EHR) data, of adults hospitalized on the hospitalist service at Yale-New Haven Hospital from 2015-2018. MAIN OUTCOME AND MEASURES: We defined hospitalists' workload as the number of patients they cared for on the first full hospital day of a given patient's encounter. We used multilevel Poisson and logistic regression to examine associations between workload and length of stay (LOS), return to the Emergency Department (ED), and readmission. We adjusted for sociodemographic factors, patient complexity and severity of illness, and weekend admission (for LOS) or discharge (for ED visits or readmission). RESULTS: We analyzed 38,141 hospitalizations. Median patient age was 64 years (IQR 51-78 years), 53% were female, and 34% were nonwhite. Mean workload was 15 patients (SD 3 patients; range 10-34 patients). LOS was prolonged by 0.05 days (95% CI 0.02, 0.08; p(0.001) when comparing the 75th workload percentile (16 patients) to the 25th workload percentile (13 patients). There were no associations between workload and ED visits or readmission within 7 and 30 days. CONCLUSIONS: There was a statistically significant but modest relationship between workload and LOS; workload was not associated with ED visits or readmissions.Given clinical reports of the deleterious effects of increased hospitalist workload, there is a need for prospective research assessing a range of outcomes, beyond those measurable in contemporary EHR data.
Authors: Stephanie K Mueller; Stephanie A Call; Furman S McDonald; Andrew J Halvorsen; Jeffrey L Schnipper; LeRoi S Hicks Journal: Am J Med Date: 2012-01 Impact factor: 4.965
Authors: Mary E Anderson; Jeffrey J Glasheen; Debra Anoff; Read Pierce; Roberta Capp; Christine D Jones Journal: J Hosp Med Date: 2015-06-30 Impact factor: 2.960
Authors: Jennifer Meddings; Heidi Reichert; Shawna N Smith; Theodore J Iwashyna; Kenneth M Langa; Timothy P Hofer; Laurence F McMahon Journal: J Gen Intern Med Date: 2016-11-15 Impact factor: 5.128
Authors: Jack Needleman; Peter Buerhaus; V Shane Pankratz; Cynthia L Leibson; Susanna R Stevens; Marcelline Harris Journal: N Engl J Med Date: 2011-03-17 Impact factor: 91.245
Authors: Hude Quan; Vijaya Sundararajan; Patricia Halfon; Andrew Fong; Bernard Burnand; Jean-Christophe Luthi; L Duncan Saunders; Cynthia A Beck; Thomas E Feasby; William A Ghali Journal: Med Care Date: 2005-11 Impact factor: 2.983
Authors: Joseph S Ross; Gregory K Mulvey; Brett Stauffer; Vishnu Patlolla; Susannah M Bernheim; Patricia S Keenan; Harlan M Krumholz Journal: Arch Intern Med Date: 2008-07-14
Authors: Kevin J O'Leary; Jonathan Turner; Nicholas Christensen; Madeleine Ma; Jungwha Lee; Mark V Williams; Luke O Hansen Journal: J Hosp Med Date: 2014-12-19 Impact factor: 2.960